NAVAL

POSTGRADUATE SCHOOL

MONTEREY, CALIFORNIA

THESIS

Approved for public release; distribution is unlimited

ENDING AMERICA’S ENERGY INSECURITY: HOW ELECTRIC VEHICLES CAN DRIVE THE

SOLUTION TO ENERGY INDEPENDENCE

by

Fredrick Stein

December 2011

Thesis Advisor: Tom Mackin Second Reader: Ted Lewis

THIS PAGE INTENTIONALLY LEFT BLANK

i

REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704–0188 Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instruction, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202–4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704–0188) Washington DC 20503. 1. AGENCY USE ONLY (Leave blank)

2. REPORT DATE December 2011

3. REPORT TYPE AND DATES COVERED Master’s Thesis

4. TITLE AND SUBTITLE Ending America’s Energy Insecurity: How Electric Vehicles Can Drive the Solution to Energy Independence

5. FUNDING NUMBERS

6. AUTHOR(S) Fredrick Stein 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

Naval Postgraduate School Monterey, CA 93943–5000

8. PERFORMING ORGANIZATION REPORT NUMBER

9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES) N/A

10. SPONSORING/MONITORING AGENCY REPORT NUMBER

11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. government. IRB Protocol number ____N/A_______.

12a. DISTRIBUTION / AVAILABILITY STATEMENT Approved for public release; distribution is unlimited

12b. DISTRIBUTION CODE A

13. ABSTRACT (maximum 200 words) The homeland/national security threat posed by the United States’ dependence on foreign oil has been part of the American discourse for years; yet nothing has been done. No pragmatic, realistic step-by-step plan has been pursued to end this scourge on the American people. The solution can be found in the problem. Net imports of oil account for approximately 50 percent of the oil the U.S. consumes. Likewise, 50 percent of oil consumed in the U.S. is consumed as motor gasoline. If overnight the U.S. stopped using oil to power its vehicles, if overnight drivers switched to electric vehicles, then overnight the U.S. would become energy independent. Using historical data to establish the effect of gasoline price changes on consumer vehicle choice, a predictive model has been created showing the expected switch to electric vehicles if the price of gasoline increases and the cost of electric vehicles decreases. There is a cost to energy independence: two to five dollars per gallon of retail gasoline sold. If monies raised from the tax are used to lower the price of electric vehicles, build recharge infrastructure, and dampen the regressive nature of the tax, energy independence is a few short years away. 14. SUBJECT TERMS Energy Independence, electric vehicle, homeland, national, security, oil, model, policy, gasoline, tax

15. NUMBER OF PAGES

86 16. PRICE CODE

17. SECURITY CLASSIFICATION OF REPORT

Unclassified

18. SECURITY CLASSIFICATION OF THIS PAGE

Unclassified

19. SECURITY CLASSIFICATION OF ABSTRACT

Unclassified

20. LIMITATION OF ABSTRACT

UU NSN 7540–01–280–5500 Standard Form 298 (Rev. 2–89) Prescribed by ANSI Std. 239–18

ii

THIS PAGE INTENTIONALLY LEFT BLANK

iii

Approved for public release; distribution is unlimited

ENDING AMERICA’S ENERGY INSECURITY HOW ELECTRIC VEHICLES CAN DRIVE THE SOLUTION TO ENERGY INDEPENDENCE

Fredrick Stein Transportation Security Administration, Field Counsel

B.S., Florida State University, 1996 J.D., University of Virginia, 1999

Submitted in partial fulfillment of the

requirements for the degree of

MASTER OF ARTS IN SECURITY STUDIES (HOMELAND SECURITY AND DEFENSE)

from the

NAVAL POSTGRADUATE SCHOOL December 2011

Author: Fredrick Stein

Approved by: Tom Mackin Thesis Advisor

Ted Lewis Second Reader

Daniel J. Moran Chair, Department of National Security Affairs

iv

THIS PAGE INTENTIONALLY LEFT BLANK

v

ABSTRACT

The homeland/national security threat posed by the United States’ dependence on foreign

oil has been part of the American discourse for years; yet nothing has been done. No

pragmatic, realistic step-by-step plan has been pursued to end this scourge on the

American people. The solution can be found in the problem. Net imports of oil account

for approximately 50 percent of the oil the U.S. consumes. Likewise, 50 percent of oil

consumed in the U.S. is consumed as motor gasoline. If overnight the U.S. stopped using

oil to power its vehicles, if overnight drivers switched to electric vehicles, then overnight

the U.S. would become energy independent. Using historical data to establish the effect

of gasoline price changes on consumer vehicle choice, a predictive model has been

created showing the expected switch to electric vehicles if the price of gasoline increases

and the cost of electric vehicles decreases. There is a cost to energy independence: two to

five dollars per gallon of retail gasoline sold. If monies raised from the tax are used to

lower the price of electric vehicles, build recharge infrastructure, and dampen the

regressive nature of the tax, energy independence is a few short years away.

vi

THIS PAGE INTENTIONALLY LEFT BLANK

vii

TABLE OF CONTENTS

I. INTRODUCTION........................................................................................................1 A. PROBLEM STATEMENT .............................................................................1

1. The Scope and Ramifications of Dependence on Foreign Oil ..........1 2. A Path to Independence.......................................................................2

B. RESEARCH QUESTIONS .............................................................................4 1. Primary Question .................................................................................4 2. Secondary Questions ............................................................................4

C. SIGNIFICANCE OF RESEARCH ................................................................5

II. LITERATURE REVIEW ...........................................................................................7 A. DEPENDENCE ON FOREIGN SUPPLIES OF OIL AS NATIONAL

SECURITY/ HOMELAND SECURITY ISSUE ..........................................7 1. Direct Threats to the Supply of Foreign Oil ......................................7 2. Secondary Effects of Foreign Oil Dependence ................................10

a. Dependence of Emergency Responders on Reliable Oil Supply ......................................................................................10

b. Impact of Natural Disasters ....................................................11 B. OIL SOURCES AND USES ..........................................................................12 C. ACHIEVING ENERGY DEPENDENCE ...................................................14 D. ADOPTION OF ALTERNATIVE FUEL VEHICLES ..............................15

1. Large-Scale Efforts in the 1990s .......................................................15 2. Predictions of Adoption .....................................................................15 3. Recent Stimulus Money .....................................................................16

E. SUPPLY AND DEMAND—ECONOMICS OF CHOICE ........................18 1. Elasticity of Gasoline .........................................................................18 2. Gasoline Price and Vehicle Choice ...................................................19

F. CONCLUSION ..............................................................................................20

III. METHOD ...................................................................................................................23 A. CASE STUDY CONSIDERED .....................................................................23 B. MODELING CHOSEN .................................................................................23 C. LIMITATIONS OF MODELING ................................................................24

IV. RESULTS ...................................................................................................................27 A. PREMISE OF THE PREDICTIVE MODEL .............................................27 B. DEVELOPMENT OF THE PREDICTIVE SALES MODEL ..................28 C. APPLICATION OF THE MODEL TO FORECAST SALES ..................36 D. FORMULAIC SHORTCOMINGS ..............................................................39

1. Use of Hybrid Coefficients ................................................................39 2. The Model Does not Reflect the Bounded Nature of the Market ..39 3. Inherent Risks of Prediction .............................................................40

V. CONCLUSIONS ........................................................................................................43 A. SUGGESTED COURSE OF ACTION ........................................................43

viii

1. Excise Tax on Retail Gasoline Purchases ........................................43 2. Recharge Infrastructure ....................................................................44 3. New Power Stations and Infrastructure ..........................................45

B. CRITICAL SECONDARY CONCERNS ....................................................46 1. Public Support Financially ...............................................................46 2. Public Support Politically..................................................................48 3. Source of Energy Storage Materials ................................................48

C. UNINTENDED CONSEQUENCES AND SECONDARY CONCERNS ...................................................................................................49 1. Fungible Nature of the Oil Market ...................................................49 2. A Regressive Excise Tax ....................................................................50 3. Risk Increasing Cost of Goods / Inflation ........................................51 4. Significant Drop in the Price of Oil ..................................................51

D. FOLLOW-ON RESEARCH .........................................................................52 1. A More Complex Version of the Present Model .............................52 2. A More Complex Modeling System..................................................52 3. Determine Coefficients for EVs Independently ..............................53 4. Determining the Cost of Energy Dependence .................................53 5. Infrastructure Questions ...................................................................53

E. CONCLUSION ..............................................................................................54

LIST OF REFERENCES ......................................................................................................55

INITIAL DISTRIBUTION LIST .........................................................................................67

ix

LIST OF FIGURES

Figure 1. Oil Imports, Consumption and Amounts Used for Motor Gasoline (After EIA, 2010a, pp. 128, 148) ..................................................................................3

Figure 2. Uses of Oil (EIA, 2010a, p. 148) .....................................................................13 Figure 3. Recharge Levels and Time to Charge (From Dickerman, 2010) .....................17 Figure 4. Comparison of the Variation of ln(Shy) with Year of Sale as Derived by

Model (blue) with Historic Data (red). ............................................................35 Figure 5. Comparison of the Variation of ln(SSUV) with Year of Sale as Derived by

Model (blue) with Historic Data (red) .............................................................35 Figure 6. Predicted EV Sales for the Period 2011–2018 in Response to Taxation and

Subsidy Policies as Projected by the Model ....................................................38

x

THIS PAGE INTENTIONALLY LEFT BLANK

xi

LIST OF TABLES

Table 1. Historical Data for SUVs and Hybrid Vehicles ...............................................33 Table 2. Coefficients of the Sales Volume Models (Eqns. 7 and 8) of Hybrid and

SUV Vehicles as Derived by Fitting These Equations to Historical Sales Data (Table 1) ..................................................................................................35

Table 3. Excise Tax Collected .......................................................................................39

xii

THIS PAGE INTENTIONALLY LEFT BLANK

xiii

LIST OF ACRONYMS AND ABBREVIATIONS

ASCM Anti-Ship Cruise Missile

CBO Congressional Budget Office

CAFE Corporate Average Fuel Economy

DOE United States Department of Energy

EV Electric Vehicle

EIA United States Energy Information Administration

FTC Federal Trade Commission

IAFC International Association of Fire Chiefs

MSRP Manufacturer’s Suggested Retail Price

PHEV Plug-in Hybrid Electric Vehicle

TEPCO Tokyo Electric Power Company

xiv

THIS PAGE INTENTIONALLY LEFT BLANK

xv

ACKNOWLEDGMENTS

To my wife, Tammar, all my love and thanks. You supported me from the

moment I mentioned the master’s program; you supported the family through all my trips

to Shepherdstown and during all my disappearing acts on the weekend. Had you simply

agreed to the sacrifices of the program, it would have been enough. Had you merely sent

me off with your love and the knowledge that our family was well during my absences, it

would have been enough. Yet you encouraged me when I faltered, advised me when I

deliberated and led the celebration as I succeeded.

To my father-in-law Gaby Laufer, I do not exaggerate when I say that without

your assistance this particular thesis would not exist. I had an idea and concept but

lacked the means to give it true academic rigor. The predictive model in this thesis bears

your sweat, your brilliance, and perhaps most importantly, your patience. You will

always have my gratitude and love.

Rarely does anyone achieve anything really worthwhile alone. That was never

truer than with this thesis and the master’s degree itself. Together with my classmates

and instructors, NPS CHDS Cohorts 1003/1004, I was never alone. The experience and

wisdom of the group was humbling. The willingness of everyone to give of themselves

in aid of all our classmates was inspiring. This thesis was only possible because I learned

to think differently and to view the world from a different perspective. That profound

insight did not come from any of our readings, papers or quizzes. It came from all of

you, my CHDS family.

At times it seemed as though this particular thesis was never meant to be. To

borrow from General McAuliffe, my interpretation of Dr. Bellavita’s response to all the

naysayers can be summed up in one word. Nuts! Dr. Bellavita’s belief in my ability to

write this thesis was contagious.

Months of research and a multitude of false starts had left me with grand ideas

and a mountain of papers, but nothing resembling a thesis. I had lost sight of the forest.

Dr. Pelfrey, Jr. has a singular ability to take something complex and not only make it

xvi

simple, but make it simple for everyone else. With his invaluable assistance, my ideas

and my mountain of research fit together almost effortlessly. Instead of a mass of

unorganized data, ideas and citations, I have a thesis. Thank you, Dr. Pelfrey.

The idea of simultaneously working through my classes, writing a thesis and

attending classes while keeping up with my work at the Transportation Security

Administration seemed impossible. If not for the support of my Office of Chief Counsel

brethren, in particular, Brandon Straus, Ron Kilroy and Richard DiGiacomo it would

have been impossible. I am in your debt.

No paper becomes a thesis without the direction, support and wisdom of thesis

advisors, and this thesis is no different. Tom Mackin and Ted Lewis not only guided

prodded and forced me to turn over stones I might have chosen to leave unturned, but

they did it with excitement and verve. I believe I am better off for it, and I know my

thesis is stronger and deeper thanks to their unwavering guidance.

xvii

THIS PAGE INTENTIONALLY LEFT BLANK

1

I. INTRODUCTION

A. PROBLEM STATEMENT

Dependence on foreign oil presents a growing threat to U.S. national and

homeland security. This threat manifests itself in several ways: it places unhealthy

restraints on U.S. and allied nations’ policy choices; it weakens the nation economically

by adding to the trade deficit; it forces our military to protect vital oil trade routes; and it

strengthens our adversaries by providing funding for their activities. Given that the

amount of oil imported by the country is roughly equal to the amount of oil used in the

U.S. to power automobile transportation, a shift to alternative transportation technologies

could dramatically improve our national and homeland security. Electric vehicles (EVs)

provide the most promising path toward ending our dependence on foreign oil.

For purposes of this thesis, “dependence on foreign oil” is the situation where the

domestic demand for oil exceeds the available domestic supply. When the domestic

supply is insufficient, some domestic consumption of oil must be satisfied by sources that

originate in locations other than the states, territories and possessions of the United

States. With this definition in mind, net importation of oil is more relevant than total

importation numbers though much of the literature relies on total importation data.

1. The Scope and Ramifications of Dependence on Foreign Oil

In 1973/1974, the U.S. imported approximately 28 percent of the oil that it

consumed (U.S. Energy Information Administration [EIA], 1998). At only 28 percent,

the six-month long 1973/1974 oil embargo caused gas prices to triple and long lines at

gas stations (Leotta, 2006, p. 4). Today, the United States imports approximately 50

percent of the oil that it consumes (EIA, 2011d). Imagine what would happen should a

similar disruption occur, now that the degree of dependence on foreign oil has markedly

increased.

The economic cost of dependence on foreign oil is staggering. The United States

has an oil trade deficit of approximately $1,000,000,000 per day, larger than our trade

deficit with China (Powers, 2010), which in 2010 was approximately $748,000,000 per

2

day (U. S. Census, 2011). Oil consumption represents 40 percent of America’s energy

needs, with 20 percent of the oil the U.S. consumes coming from the Persian Gulf Region

(Cohen, 2007). The cost to the United States is compounded. Not only do we send one

billion dollars out of our economy every day, but much of that same money is then used

in a manner that directly threatens our security. Time and again, the U.S. military and

national security leaders have warned of the substantial risk this outflow of capital poses

to the security of the United States. For example, Vice Admiral Dennis McGinn has

cautioned that the oil trade deficit, much of it enriching nations that wish us harm, is an

unsustainable transfer of wealth that has us literally funding both sides of the conflict in

the “war on terror” (Powers, 2010). Former national security adviser Robert McFarlane

and former CIA director R. James Woolsey, have recently described our dependence on

foreign oil as, “the well from which our enemies draw their political strength and

financial power: the strategic importance of oil, which provides the wherewithal for a

generational war against us” (McFarlane, 2011). When asked about the most important

area of research for aiding those under his command, Marine General James Mattis

responded by imploring that the country, “Unleash us from the tether of fuel” (Powers,

2010). Similarly, a panel of 11 former generals and admirals testified that “dependence

on foreign oil leaves us more vulnerable to hostile regimes and terrorists” (Powers,

2010).

The restraints from oil dependence on U.S. foreign policy decisions are untenable.

For example, as the nation struggles to blunt Iran’s drive towards nuclear weapons,

administration officials are wary of enacting sanctions on Iranian oil exports. They fear

that it might drive up oil prices when the United States and European economies are

weak. As they consider the probable impacts, they are forced to conclude that in the end,

they simply cannot predict what would happen (Sanger, 2011).

2. A Path to Independence

Energy dependence is a product of the United States’ importation of oil. Though

the U.S. is the third largest producer of oil in the world (EIA, 2010b; EIA, 2011d), it still

imports roughly 60 percent of the total oil it consumes (Brooker, 2010). Since the U.S.

3

also exports approximately two million barrels of oil per day, net imports are

approximately 50 percent of U.S. consumption (EIA, 2010a, p. 128). Roughly two-thirds

of the oil used in the United States is for transportation (Etezadi-Amoli, 2010). Virtually

none of the oil consumed (approximately 0.2 percent in 2009) in the United States is used

for electricity production (EIA, 2010a, p. 152). That means that adding nuclear or clean

coal facilities, building wind farms, installing solar panel fields, etc., do little to foster

energy independence. Those technologies may replace fossil fuel combustion for the

generation of electricity but not in a manner that can currently be utilized by most of the

transportation sector that depends almost exclusively on the combustion of oil.

Electric vehicles can change all that. They provide a path for eliminating our

dependence on foreign oil. There is a close, though not perfect, equivalence between the

amount of net imported oil, and the amount of oil used for motor gasoline. Of the 18.7

million barrels of oil consumed per day in the U.S., approximately nine million barrels

are used as motor gasoline (EIA, 2010a, p. 148). Motor gasoline is by far the single

greatest use of oil (see Figure 1). Replacing the other uses of oil would require a half

dozen or more other changes, some of which are not yet technically feasible. EVs shift

the energy required for automobile transportation to the electric power sector, and, as

long as that electric power is generated using domestically available sources of energy

other than oil, the U.S. would reduce its overall consumption of oil by roughly 50

percent. This establishes that if the United States were to cease using oil to power

automobiles, it would effectively eliminate its dependence on foreign oil.

Figure 1. Oil Imports, Consumption and Amounts Used for Motor Gasoline (After EIA,

2010a, pp. 128, 148)

4

There is analysis of recent historical fluctuations in gasoline prices that looks

beyond simple supply and demand. It shows that with long-term price increases in the

cost of gasoline, people will reduce their consumption of gasoline, at least partially, by

moving to more fuel-efficient vehicles (Congressional, 2008).

EVs provide the best means of eliminating the use of petroleum products to power

automobiles in the United States. By moving to EVs (provided the increased electricity

demands are not met through the use of petroleum), the yellow bar in Figure 1 can be

eliminated, ending the need for the oil represented by the red bar, net imports of oil. It is

important to note, however, that this transformation will require upgrades to and

expansion of the nation’s electric power infrastructure, thereby requiring infrastructure

investments. More details on that issue are discussed in Chapter V.

Although the move to EVs is likely to occur slowly in response to market forces

alone (Hensley, 2009), given the urgency of the security threat posed by dependence on

foreign oil, there is a dearth of policy designed to increase the pace of transformation to

an EV fleet. While there are many discussions of the effect of EVs on the environment,

there are no policies that look to speed up this transformation from a national or

homeland security perspective. Speeding up EV adoption depends on several factors,

including: improving battery efficiencies and capacities to increase range; bringing the

total cost of EV ownership in line with or below the total cost of gasoline powered

vehicles; increasing the pace of EV infrastructure development (e.g., recharge/refuel

facilities); and increasing consumer adoption of EVs (Hensley, 2009).

B. RESEARCH QUESTIONS

1. Primary Question

How can various policy choices affect consumer adoption of electric vehicles to

move the United States from internal combustion vehicles to electric vehicles?

2. Secondary Questions

1. Which policy choices would lower the demand for gasoline powered automobiles?

5

2. Which policy choices would raise demand for electric vehicles? Although this question is related to secondary question 1, there may well be policies that lower the demand for internal combustion vehicles without specifically increasing the demand for EVs.

3. What are the potential unintended consequences of policy choices favoring electric vehicle adoption?

C. SIGNIFICANCE OF RESEARCH

This thesis will fill the void in the literature looking at EV adoption as a

national/homeland security concern. There is ample analysis of EV adoption from an

environmental and purely economic perspective, but there seems to be no literature that

gives serious consideration to adopting EVs as a potential solution to the security

vulnerability created by energy dependence. Additionally, there does not appear to be a

comprehensive look at how various policy choices might be used specifically to bring

about the speedy adoption of EVs. This thesis will assimilate the available literature and

data to establish (1) that EVs have the potential to solve the security vulnerability

inherent in energy dependence and, (2) what policies can be implemented to bring about

the speedier adoption of EVs in the United States.

6

THIS PAGE INTENTIONALLY LEFT BLANK

7

II. LITERATURE REVIEW

Oil is often the topic of public debate, political discourse and academic research.

A great deal of time and resources are spent discussing the sources of oil, the

environmental and geopolitical effects of oil and dependence on foreign sources for oil

and how the nation might reduce its use of oil/foreign oil. There is also literature

discussing the national/homeland security implications of energy dependence; however,

there is an absence of literature either discussing how EV adoption might solve that

security threat or describing what policy choices can be used to speed up EV adoption.

A. DEPENDENCE ON FOREIGN SUPPLIES OF OIL AS NATIONAL SECURITY/ HOMELAND SECURITY ISSUE

1. Direct Threats to the Supply of Foreign Oil

There is ample material establishing the existence of the security threat posed by

dependence upon foreign sources of oil. At the surface, rhetorical level, there exists a

robust record of high visibility proclamations that energy dependence is a critical security

issue for the United States (for example, see CNN, 2010). Additionally, for decades the

nation’s leaders have recognized the security importance of energy independence,

espousing its centrality to national security. Video clips of every U.S. president from

Richard Nixon to the current President capture each of them discussing the importance of

ending the United States’ dependence on foreign oil (Stewart, 2010). There is also ample

media coverage of the rhetoric of nations hostile towards the U.S. describing how they

are willing to cut off the United States’ supply of oil. This is primarily centered on oil

producing countries from the Middle East and Venezuela. This literature tends to be

sound bites without much meat, but the advantage is that it is designed for public

consumption, helping to socialize the idea of energy dependence. President Obama

recently announced that for “geopolitical and economic reasons,” the nation needed to

reduce its reliance on imported oil by about 30 percent over the next 10 years (Broder,

2011).

8

A more in depth body of literature examining energy dependence also exists. The

1973/1974 oil embargo provides historical precedence for foreign nations using oil as a

lever to affect U.S. national policy (Paust, 1974). In 1973/1974, the U.S. imported

approximately 28 percent of the oil that it consumed (EIA, 1998). The embargo,

sometimes referred to as “Energy Pearl Harbor Day” (Light, 1976), was sufficiently

severe to serve as the catalyst for legislative action concerning the use of fuel. Gas

shortages were significant enough that, along with a tripling of world oil prices, gasoline

consumption in the U.S. actually decreased by 2 percent in response to the embargo

(Leotta, 2006, p. 4). The situation was so severe that naval oil reserves were tapped for

emergency civilian supplies (Light, 1976).

Literature has addressed the present threat from state actors. Hugo Chavez,

president of Venezuela, has threatened to cut the supply of oil to the U.S., not because of

the threat of a U.S. invasion, but as leverage to prevent Colombia from invading

Venezuela (CNN, 2010). Similarly, Russia has shown it may be willing to take military

action to control the supply of oil flowing to the West. The invasion of Georgia may

have been more about the Baku-Tbilisi-Ceyhan pipeline (a conduit of oil to the West

owned by U.S. and British energy firms) than about support of the separatists in South

Ossetia (Kimery, 2008).

Similar threats have come from non-state actors like Osama bin Laden and

Ayman al-Zawahiri, who have called for attacks on economic assets, especially energy

sources (Cohen, 2007). In a tape aired by Al-Jazeera in February 2006, Zawahiri said, “I

call on the mujahedeen to concentrate their attacks on Muslims’ stolen oil, most of the

revenues of which go to the enemies of Islam while most of what they leave is seized by

the thieves who rule our countries” (Cohen, 2007). The non-state threats are broader than

just Al Qaeda with groups specifically attempting to wreak havoc in international markets

(Worth, 2010).

These attacks have real consequences, and the world oil market recognizes the

risk posed by these attacks. In February of 2005, a failed Al Qaeda attack on the Aramco

facility in Abqaiq, Saudi Arabia, caused the price of oil on international markets to jump

nearly two dollars per barrel (Cohen, 2007).

9

As grave as those threats are, perhaps the single greatest threat comes from Iran.

The literature on this subject includes Congressional investigation of the risk. Some

members of Congress feel that in response to any attack, Iran would retaliate by

attempting to shut off the supplies of Middle East oil (Chairman, 2007). Those feelings

are supported both by the public expressions of intent from Iranian officials, and by the

capabilities of the Iranian military (Berman 2006; Cohen 2007). These are not idle

threats. If Iran retaliated and shut down the Strait of Hormuz, it would mean the

temporary loss of more than 15 million barrels of oil a day (Cohen 2007). Cohen goes on

to describe just how effectively Iran has built a military arsenal with this capability in

mind.

Iran boasts an arsenal of Iranian-built missiles based on Russian and Chinese designs that are difficult to counter both before and after launch. Of particular concern are reports that Iran has purchased the SS-N-22 Moskit/Sunburn anti-ship missile. The supersonic Sunburn is specifically designed “to reduce the target’s time to deploy self- defense weapons” and “to strike ships with the Aegis command and weapon control system and the SM-2 surface-to-air missile.” 12 Iran is also well stocked with older Chinese HY-1 Seersucker and HY-2 Silkworm missiles and the more modern C-802 anti-ship cruise missile (ASCM)—designs that Iran has successfully adapted into their own Ra’ad ad Noor ASCMs.

Iran has a large supply of anti-ship mines, including modern mines that are far superior to the simple World War I–style contact mines that Iran used in the 1980s. They include the Chinese- designed EM-52 “rocket” mine, which remains stationary on the sea floor and fires a homing rocket when a ship passes overhead. In the deep waters in the Strait of Hormuz, such a weapon could destroy ships entering or exiting the Persian Gulf. According to one expert, Iran “can deploy mines or torpedoes from its Kilo-class submarines, which would be effectively immune to detection when running silent and remaining stationary on a shallow bottom justoutsidetheStraitofHormuz.”13 Iran could also deploy mines by helicopter or small boats disguised as fishing vessels.

Mines are only one of a host of potential Iranian threats to shipping in the Persian Gulf. The naval commandos of Iran’s Revolutionary Guards are trained to attack using fast attack boats, mini-sub- marines, and even jet skis. The Revolutionary Guards also have underwater demolition teams that are trained to attack offshore oil platforms and other facilities. (Cohen 2007)

10

Even with the United States’ extensive military power, U.S. intelligence estimates

that Iran’s military rearmament has given it the ability to shut off the flow of oil from the

Persian Gulf temporarily (Berman, 2006).

Literature has examined the fungible nature of oil as an asset and how that

increases the threat to the United States. Although the United States imports much of its

foreign oil from friendly nations like Canada and Mexico (EIA, 2011a), the entire world

supply would be impacted by a significant disruption anywhere in the oil market

(Chairman, 2007). Should Iran or another actor temporarily interrupt the world wide flow

of oil, if the United States chooses to respond, for example to open the Straits of Hormuz,

the government will be forced to make very difficult choices. According to Thomas,

“Reserve stores of petroleum and petroleum-based fuels would dwindle quickly—

particularly during wartime operations—leaving the U.S. military unable to obtain

suitable alternative fuels and rendering it virtually immobile” (2010).

2. Secondary Effects of Foreign Oil Dependence

This body of literature tends to focus on the importance of oil in general. There is

no real discussion of the distinction between foreign and domestic oil supplies. It is

necessary to consider this body of literature, in concert with the prior literature, in order

to understand the secondary ramifications posed by dependence on foreign oil.

a. Dependence of Emergency Responders on Reliable Oil Supply

Of course, if the military is demanding all available petroleum resources,

that leaves fire, police and ambulance services in short supply in the United States.

Other, non-critical, demands for petroleum would be left completely unfulfilled (e.g.,

commuters, truckers and the airline industry). Emergency services in the U.S. are

delivered via fire trucks, ambulances and police cars; all of those vehicles depend on

access to gasoline or diesel fuel. So delivery of emergency services in the United States

depends on oil. Additionally, during catastrophic events, when emergency services are

most in need, the services’ reliance on gasoline or diesel fuel increases significantly as

they turn to generators to supply power in order to provide many of their services.

11

Disruptions are inevitable for any number of reasons, including geopolitical or terrorist

attacks on critical energy infrastructure (Leotta, 2006).

On August 14, 2003, over 9,300 square miles, covering eight states and

portions of Canada lost electrical power with virtually no warning (International

Association of Fire Chiefs [IAFC], 2003). In New York City alone, emergency services

responded to 91,000 9–1–1 calls during the outage (IAFC, 2003). Even in their reduced

capacity, emergency services were called on to perform more than 30 distinct tasks: inter

alia, elevator rescue, subway rescue, fire suppression, hazard calls, traffic accidents,

welfare checks at hospitals, senior citizen homes, day care centers and prisons, providing

power to critical care facilities, distributing water, opening emergency shelters for the

elderly and helping the elderly up and down stairs (IAFC, 2003).

The firefighters were able to perform these actions because they had fuel

in their vehicles and because most of their 911 dispatch centers and many fire stations

had emergency power (IAFC, 2003). Of course, oil fueled those backup power systems.

Emergency generators were the only means of pumping water to fight fires in some areas.

Many stations relied on mutual aid agreements. Cleveland had an almost total loss of

water, but by activating a statewide assistance plan, tankers and personnel saw them

through (IAFC, 2003). Should the crises impact the U.S. broadly, this reciprocation of

coverage would not be possible, incapacitating many first responders.

b. Impact of Natural Disasters

The effects of Hurricanes Katrina and Rita on the regionally impacted

populations have been well documented. The second order effects further establish the

homeland security implications of reliance on oil. The effects of these storms went far

beyond the areas where the storm itself hit (Leotta, 2006). The storms showed the

vulnerability of the oil supply chain well beyond the coasts directly hit by the storms.

Nearly 1.6 million barrels a day of crude oil (between seven and eight percent of U.S. oil

consumption) was produced by Gulf of Mexico federal offshore sites before the

hurricanes struck (Leotta, 2006). Production was virtually halted in the wake of both

storms, as production facilities were evacuated and wells were closed. By September 1,

12

at least one county in North Carolina was faced with 60 percent of its gas stations out of

fuel (Leotta, 2006). The shortage was occurring hundreds of miles from the storm.

Though the energy crisis was severe, it was not expected to last long. Nonetheless, it

became necessary for the state energy office to convene all major agencies and require

each to prepare a list of how they might curtail fuel consumption; agencies had to

prioritize their fuel consuming activities (Leotta, 2006). The situation was sufficiently

short lived to avoid a state of emergency.

Had the fuel shortages persisted or worsened, the state would have been

unable to provide the National Guard, who obtains its fuel through the state, with fuel for

very long (Leotta, 2006). North Carolina was fortunate that it did not need to rely on the

services of its National Guard during this crisis. One of the proposed solutions was to

have first responders and other critical users enter into firm contracts for fuel (Leotta,

2006). Unfortunately, such an arrangement only works if there is sufficient fuel. If fuel

supplies were limited, states would be forced to choose which of their vital operations,

which critical needs of their populations, to forego.

While this body of literature looks at unintended disruptions to the supply

of oil (e.g., hurricanes and accidental blackouts) there is no reason to believe the impacts

would be diminished if the disruptions were intentional. Moreover, though not discussed

in any literature, it is not difficult to imagine the magnification of devastation were an

intentional disruption of the foreign supply of oil to coincide with a natural disaster.

B. OIL SOURCES AND USES

There exists a robust cataloging and compiling of both the various uses of oil in

the United States as well as a breakdown of the sources of the oil used in United States.

The U.S. Energy Information Administration (EIA) collects, organizes and makes

available to the general public a cornucopia of data relating to oil consumption and

production. All manner of data is tracked, including total oil used, oil imports, OPEC

versus non-OPEC imports, by country OPEC imports, etc. For 2009, EIA data shows

that the United States had net imports of 9.7 million barrels of crude oil per day (EIA,

2010a). The 9.7 million barrels of crude oil imported per day was from a total of 18.7

13

million barrels of crude oil used per day (EIA, 2010a). During the same year, 13.3

million barrels of oil were used per day for transportation purposes in the United States,

with nine million barrels per day specifically used for motor gasoline (EIA, 2010a). The

nine million barrels used every day for motor gasoline represented 48 percent of all U.S.

petroleum consumption (EIA, 2009). The relative numbers have held fairly constant. In

2010, the U.S. consumed 19.1 million barrels of oil per day (EIA, 2011d), with nine

million barrels per day used for motor gasoline (EIA, 2011e). In 2010, net imports

accounted for 49 percent of U.S. oil consumption (EIA, 2011d). For a more detailed

description of the uses of oil, see Figure 2.

1 Liquefied petroleum gases. 2 Asphalt and road oil, aviation gasoline, kerosene, lubricants, naphtha-type jet fuel, pentanes plus, petrochemical feedstocks, special naphthas, still gas (refinery gas), waxes, miscellaneous products, and crude oil burned as fuel.

Figure 2. Uses of Oil (EIA, 2010a, p. 148)

There is no significant body of literature challenging the veracity of the data

provided by the EIA. There is, however, secondary literature that seeks to analyze the

available data. For example, there is literature that focuses on comparing the amount of

oil imported by the United States with its overall use of oil. In writing for the Truman

National Security Project, Jonathan Powers (2010) uses the statistics provided by EIA to

demonstrate the breadth of the national security hazards from dependence on fossil fuels,

most specifically oil. Other authors have utilized the same EIA data to demonstrate the

need and benefits that could result from switching to plug-in hybrid electric vehicles

14

(PHEV) (Brooker, 2010). Though not always with attribution, the media seems also to

rely on the data provided by EIA. The Wall Street Journal ran an article discussing why

an alternative-energy future may be far away and explained that one cause was the

amount of fuel used for transportation1 (Totty, 2010). The literature describing the

sources and uses of oil in the United States is relatively complete and settled.

C. ACHIEVING ENERGY DEPENDENCE

Notwithstanding the difficulties created by dependence on foreign sources of oil,

illustrated by history and decried repeatedly by political and military leaders, there has

been little literature focused on achieving independence from a security perspective.

There are some signs that might be changing. Recent practitioner based scholarly work

has emerged that looks directly at the problem of ending U.S. dependence on foreign oil

or oil outside of North American sources (Cowden, 2008; Thomas, 2010). These papers

describe very broad frameworks for developing energy independence, but they do not

discuss nut and bolt practicalities of how the country might arrive there from its present

state.

There are also environmentally focused papers that discuss “green energy,” that

give passing reference to the security implications in order to bolster the strength of their

arguments (for example see Hensley, 2009). Some of the “green energy” literature

describes the various policy considerations as they relate to exploring and/or encouraging

alternative sources of energy for vehicles (Whoriskey, 2009). Whether vehicle specific,

or energy use in general, the solutions involving “green energy” generally embrace local

sources of energy, consequentially reducing dependence on foreign sources of energy.

There is also analysis of policies concerning corn based ethanol (Hauser, 2007),

geothermal energy (Sweet 2010) and electric vehicle infrastructure (Cote´, 2010).

However, there is no real move in this literature to embrace security as a driver of change

and to include that focus in their analysis. Therefore, the authors fail to leverage a view

1 Totty refers to 3.3 billion barrels of gasoline used for transportation in 2009. That is not accurate, but would be consistent with the nine million barrels of oil used for motor gasoline (13.3 million barrels for total transportation) in the U.S. per day. At 365 days per year, that would roughly equate to 3.3 billion barrels of oil used for motor gasoline in 2009 per USEIA (EIA, 2010).

15

that could provide significant sources of political will. Furthermore, the literature that

focuses on energy production largely misses the mark. Since only about 0.2 percent of

the oil used in the United States is used for electricity production (EIA, 2010a, p. 152),

absent some manner of translating that energy into a form usable by the transportation

sector “green” sources of electricity production will not displace oil.

D. ADOPTION OF ALTERNATIVE FUEL VEHICLES

1. Large-Scale Efforts in the 1990s

There have been efforts in the past to bring about the adoption of EVs or other

zero-emissions vehicles. There is some literature available that looks at and analyzes

these past efforts. For example, in the 1990s, California, in a somewhat adversarial

process, enacted a series of policies, including the requirement that a small percentage of

new vehicles sold in California be zero-emissions vehicles (Calef, 2007). As time

progressed towards the deadlines set by California, it became apparent that the industry

was not going to meet the standards and therefore California shifted to an emphasis on

low-emissions vehicles (Boyd, 1996). Also in the 1990s, with little public discussion,

France engaged in a nonadversarial effort at bringing about the adoption of EVs (Calef,

2007). For a range of reasons, including technology that was not yet ready, the efforts

failed in both France (Calef, 2007) and in California (Bedsworth, 2007; Calef, 2007).

This body of literature is helpful in that it documents potential pitfalls. However,

especially because technology plays an integral role in the analysis, these case studies are

somewhat out of date.

2. Predictions of Adoption

The literature is replete with predictions of when EVs will or should encompass a

certain percentage of the automobile market share. Al Gore concluded that it is possible

to completely eliminate the internal combustion engine over a 10-year period if a

concerted effort is made to do so (Gore, 1992). Most analysis focuses on a given market

share based on the cost of oil or based on the cost of the EV batteries. One study

determined that lithium ion batteries will drop in price from $720/kWh to between

$405/kWh and $445/kWh by 2020, resulting $510 million in sales of EVs if oil trades at

16

$70 per barrel and $9 billion in EV sales if oil trades at $200 per barrel (Haldis, 2009).

Another, more optimistic analysis by automotive market research company CSM,

predicted that by 2020, battery prices will cost only 20 percent of what they cost in 2009

(Hammerschmidt, 2009). Another study by UC Berkely predicted that EVs could make

up as much as 64 percent of the new-car market share by 2030 (Peckham, 2009).

There has even been discussion of what it might take to meet President Obama’s

target of one million EVs on the road by 2015 (Daigneau, 2011). A study by Credit

Suisse, a financial services company, anticipates a faster development of the market,

predicting that there will be five million EVs on the road by 2020 (Sherman, 2011). This

body of literature offers predictions of where the country might end up with some general

explanations of the factors that impact which future materializes, but it fails to provide

sufficient analysis of the interplay of those factors and does not adequately describe how

those factors can be utilized purposefully to achieve the wide scale adoption of EVs.

3. Recent Stimulus Money

One effort at increasing the market share of EVs was the recent stimulus package;

$2.4 billion in Recovery and Investment Act money was dedicated to advances in EV

batteries, building the domestic EV industry, and building charging stations2 (Lee, 2010).

Data has been collected capturing the amount of investment and in some cases the

specific returns or anticipated returns.

One area that has received significant attention is recharge infrastructure. Though

EVs can be charged at home, many EVs currently have limited range compared to their

internal combustion counterparts. For example, the Nissan Leaf can travel approximately

73 miles3 on a single charge (DOE, 2011a), though some models of the Tesla Roadster

can travel 300 miles on a single charge (Tesla, n.d.). The phenomena known as “range

anxiety” can diminish utilization of EVs, with use well below the actual range of the EV,

2 These funds are separate from the funds used to give $7,500 in tax rebates to customers who purchase EVs (U.S Department of Energy [DOE], 2011a).

3 It is not uncommon to see articles discussing the Leaf describe it as achieving 100 miles to the charge. Though under certain conditions, Nissan claims this to be true, what is probably being cited is the miles per gallon equivalency of the Leaf, which has been calculated to be the equivalent of 99 mpg (DOE, 2011a).

17

making a recharge infrastructure critical (Botsford, 2009). There are various types of

recharge facilities. The cost of the slow charge facilities is significantly cheaper than for

fast charge. In fact, some EVs can slow charge without any change to a standard

electrical outlet (Dickerman, 2010; German, 2010). As shown in Figure 3, as the speed of

the charge goes up, so does the required voltage/amperage output, which necessitates

more expensive facilities.

Figure 3. Recharge Levels and Time to Charge (From Dickerman, 2010)

One estimate of the cost4 of recharging stations can be taken from a project

underway in Portland, Oregon. Oregon is using a two million dollar federal grant

(stimulus money) to build 42 “quick charge” stations along the I-5 corridor, ensuring no

gap greater than 50 miles (Oregon, 2010). That averages out to $47,619 federal dollars

per station. These stations are designed to charge an EV battery to 80 percent capacity in

a 20 to 30 minute period (Oregon, 2010).

A project funded in part by the U.S. Department of Energy (DOE) provides an

estimate for the cost of building slow charge stations. Given the total cost of $230

million (of federal money) to build 15,000 charging stations, each station costs

approximately $15,333; however, that also includes 310 quick charging stations (Read,

2010), which are more expensive. Assuming the quick charging stations cost $47,619

each, the cost per station of the ordinary charging stations would be approximately

$14,652 federal dollars.

4 The cost described here, is the cost to the federal government. In order to receive the stimulus funds,

the recipients had to match the funds dollar for dollar from private or state/local government sources (Lee, 2010). Therefore, the total cost is actually twice that represented here, but the cost to the federal government is accurately described.

18

The DOE has released its analysis of the Recovery and Investment Act stimulus

funds. In its estimation, it is on track to meet President Obama’s goal of one million EVs

by 2015 (DOE, 2011b). The scientific community has taken note of the anticipated

benefits from the stimulus fund, highlighting certain DOE predictions. One such

prediction is that a battery capable of powering a vehicle 100 miles, which costs $33,000

today, will cost just $16,000 by the end of 2013 because of those stimulus fund

investments (Stimulus, 2010).

This body of literature is fairly thin. It consists mostly of the documentation of

data, without much analysis. It is sufficient to provide a qualitative assessment of the

bang for the buck in estimating the expense of building a recharge infrastructure to

support EVs. The research does not attempt to extrapolate forces like economies of scale

that might reduce the cost of infrastructure development. It also fails to analyze whether

or not the recharge facilities can be expected to be or become economically self-

sustaining and thus whether pure commercial forces can be relied upon in the future to

build the recharge infrastructure. The creation of recharge infrastructure and an in-depth

description of range anxiety are addressed in more detail in Chapter V.

E. SUPPLY AND DEMAND—ECONOMICS OF CHOICE

1. Elasticity of Gasoline

Academics and the federal government have both examined the elasticity of

gasoline. That is, what happens to the amount of gasoline sold when the price of gasoline

goes up or down. There are a multitude of studies on this topic; however, two papers have

looked at many of the gasoline elasticity studies conducted over the past few decades in

an effort to summarize the field. These meta-studies essentially conclude that gasoline

elasticity is two pronged. There is a short-term affect that is particularly small and then a

long-term affect that is somewhat larger (Espey, 1996; Goodwin, 2003). The study by

Espey focuses on elasticity in the United States, while Goodwin’s study is focused on the

United Kingdom. Goodwin’s study also looks at the reduction in traffic volume from an

increase in gasoline price. Morris (2007) has also described this understanding of the

elasticity. Speaking at an EIA conference, he described the same two-phased elasticity.

19

In the short run, elasticity was between zero and -0.5, and in the long run, it was between

-0.5 and -1.5. Put another way, in the short run, a gasoline price increase of 10 percent

will reduce the quantity demanded by two percent. In the long run, a sustained 10

percent increase in the price of gasoline will result in a six percent decrease in

consumption (Federal Trade Commission [FTC], 2005).

The elasticity of gasoline is well documented. What this body of literature does

not address is the effect, if any, that gasoline prices have on consumer vehicle choices.

2. Gasoline Price and Vehicle Choice

Around the turn of the millennium, carbon tax and other policy discussions about

global warming brought the question of gasoline elasticity into sharp focus (Kayser,

2000). Rather than look just at more traditional measures of elasticity, the literature

evolved in an effort to account for income disparities, miles driven and, ultimately,

vehicle choice. Gasoline demand and the demand for automobiles were modeled as a

joint decision (Kayser, 2000). By also accounting for income, it emerged that there was

not a uniform elasticity measure, but rather elasticity varied across the income

distribution (Kayser 2000). Importantly, it was also determined that gasoline demand

responds to changes in the price of gasoline in large part by modifying the fuel efficiency

of the car fleet rather than through an adjustment of miles traveled (Jeihani & Sibdari,

2010; Kayser, 2000). In fact, it has been determined that the number of miles driven

bears almost no relation to gas price changes (Jeihani & Sibdari, 2010).

The federal government has also explored this topic and come to similar

conclusions. Looking at the period from 2003 through 2007, the Congressional Budget

Office (CBO) determined that the 100 percent increase in gasoline prices had induced

motorists to adjust the types of vehicles they purchased (CBO, 2008). They note that in

response to consumer demand, the fuel economy of new cars increased in nearly every

new model year since 2000, despite a slightly larger annual price increase for more fuel-

efficient vehicles (CBO, 2008). The report specifically notes the potential policy

advantages that tax increases on the price of gasoline have over increasing the federal

corporate average fuel economy (CAFE) standards. That advantage stems from more

20

direct alignment with market forces (CBO, 2008). The literature also notes that in order

for elasticity to strengthen, consumers must have substitutes that are easy to switch to

(FTC, 2005).

Another government sponsored review of vehicle electrification focused on the

different economic effects of various choices (e.g., plug-in hybrids or pure EVs),

opportunity charging versus charging at night versus dynamic charging as the vehicles

drive along stretches of road, etc (Brooker, 2010). That analysis concluded that the most

economic future lay in dynamic charging, but the study ignored infrastructure costs, used

an exceptionally conservative estimate of only a three percent reduction per year in

battery cost, and, while allowing other factors to change over time, used a fixed gasoline

price based on the average cost of gasoline in 2008 (Brooker, 2010).

As with pure elasticity, hard numbers have been developed in this body of

literature. If gas price increases by 10 percent, the demand for SUVs will decrease by

13.7 percent and the demand for hybrid cars will increase by 9.1 percent (Jeihani &

Sibdari, 2010).

This body of literature is relatively new and to some degree at risk of becoming

stale due to fast-paced changes. The availability of alternative vehicles, EVs versus

hybrids for example, quickly shifts the dynamic, especially if economies of scale are

reached. However, the demonstrated principle of changes in consumers’ choice of vehicle

adds significant value by providing a reasoned model to enhance the predictive process.

What remains absent is both a focus on the risks and costs of energy independence

(though it is often mentioned in passing) and an analysis of specific targeted policy

choices and how they might bring about change—policies that raise the price of gasoline,

lower the price of EVs, and otherwise promote the adoption of EVs.

F. CONCLUSION

Existing literature provides ample support establishing the extent of and threat

posed by the United States’ dependence on foreign oil. Current literature sufficiently

describes the various uses petroleum is put to in the United States. There is some

discussion focused on alleviating the dependence on foreign oil by encouraging

21

alternative energy and alternative energy vehicle adoption, but there is a glaring

omission. The focus of the literature is really on economic and environmental factors and

benefits. It does not look at EV adoption with an eye truly focused on U.S. security. The

existing literature also fails to use that security perspective to consider specific policy

choices that might be made to increase the pace of EV adoption in order to bring about

energy independence.

22

THIS PAGE INTENTIONALLY LEFT BLANK

23

III. METHOD

A. CASE STUDY CONSIDERED

The case study method is one potential method to assess how policy choices can

be used to bring about the shift to EVs. There is data available and have been studies

made of the efforts of California and France to move toward zero-emissions vehicles in

the 1990s (Bedsworth, 2007; Calef, 2007). There are limitations to this method that

make it less than ideal. In the first instance, the adoption of EVs is connected to the

current state of technology. California’s program had to be redirected to low emissions

vehicles rather than zero emissions, precisely because of technology limitations (Boyd,

1996). Those efforts were before the advent of the Tesla Roadster, Nissan Leaf, Chevy

Volt and other market ready EVs. The technology of the 1990s does not resemble present

day EV technology. For that reason, the case study method is not an ideal choice.

B. MODELING CHOSEN

This thesis will utilize predictive modeling to examine the effect different policy

choices would have on the adoption of an EV fleet in the United States, which would

bring about U.S. energy independence. Predictive modeling provides a mechanism for

policymakers to visualize the likely effect of policy choices as an aid in determining the

most effective combination of policy choices to bring about EV adoption and thereby

achieve energy independence.

The thesis will utilize inductive modeling based on historical socio-economic and

vehicle data. A model, based on the concept presented by Jeihani and Sibdari (2010), will

be created to predict the trend of increased sales of EVs that would result from varying

levels of increased gasoline prices and decreased EV prices. The system of vehicle

choice is extremely complex. There is no expectation that the model will produce

numbers that predict future sales with exact precision. Rather, the model is expected to be

sufficiently accurate to establish turning points. On a graph showing sales, the model

should be able to show the “hockey stick” affect accurately, the point where the rate of

sales increases dramatically. The model can then be used by anyone, inputting different

24

assumptions about the future and accounting for different policy choices intended to

speed up the adoption of EV sales. Specifically, the model will allow the policymaker to

view a predicted affect from a near limitless number of variations of gasoline excise tax

decisions that feed into a rebate, lowering the cost of EVs.

C. LIMITATIONS OF MODELING

As every mutual fund prospectus warns, past performance is no guarantee of

future success. Predictive modeling has the same inherent risk. The modeling method

utilizes past data results to predict future results. It is inevitable that there are other,

unaccounted for, contributors to the observed historical behavior. What is not known is

how significant those other forces are. It is also possible that people will simply decide to

behave differently. For example, if the issue becomes politicized and people modify their

behavior accordingly, or groups with opposing financial interests flood the discussion

with information that is either untrue or misleading, a model relying on past behaviors

may lose its efficacy. Finally, it is possible that the historical data is accurate, but only

within the ranges that have historically occurred. For example, assuming the largest

gasoline price increase accounted for in creating the model is X, when a model used to

predict the results that would occur for a gasoline price increase of 2X, it is possible the

model loses its efficacy.

Though predictive modeling entails risks, those risks are somewhat mitigated by

the breadth of the data reviewed and by the range of factors considered. The Jeihani and

Sibdari (2010) concept incorporates far more than just the change in demand resulting

from a gasoline price increase; it incorporates vehicle type and fuel efficiency, household

disposable income and unemployment rates as well. In fact, by use of a stand-alone

catchall constant, to some degree it incorporates all factors affecting choice, even where

those factors are indeterminable. By including these additional criteria, the formula

reduces the chance there is some other factor being overlooked.

25

Predictive modeling provides a tool for policymakers, giving them a logical

framework for assessing the most likely outcome of their decisions. Though not perfect,

use of historical factual data greatly improves upon gut instinct, or other haphazard

efforts at predicting the outcome of policy decisions.

26

THIS PAGE INTENTIONALLY LEFT BLANK

27

IV. RESULTS

A. PREMISE OF THE PREDICTIVE MODEL

The goal of this research is to provide a rigorous basis for federal policies that

would accelerate the nation’s transition from gasoline internal combustion engines to a

national fleet of EVs. By eliminating the demand for motor gasoline, the overall daily

demand for oil in the U.S. would approximately equate to the amount of oil the U.S.

produces each day. The primary levers of any federal policy are likely to be the price of

gasoline and the price of EVs.

In this chapter, an effort is made to model the influences on vehicle choice,

focusing on the consumer’s decision to purchase a more fuel-efficient vehicle. The

model looks at the basic costs of vehicle ownership as well as the more complex factors

of vehicle choice. The methodology will provide a model that is accurate at lower

volumes of sales, recognizing that given the overall complexities of the system, it cannot

predict precise sales numbers. The model will also create a general methodology where

the practitioner can incorporate additional factors, or utilize different assumptions, and

have a grounded basis to predict the results flowing from those choices.

The model developed by Jeihani and Sibdari (2010) provides the conceptual basis

for development of a predictive model. They utilize a binary vehicle choice model in

order to provide a quantitative framework to assess the various factors that might

influence consumer transition choices (Jeihani & Sibdari, 2010). In their work, the

relative probability, P, of choosing one type of car, denoted by the subscript e, to another

type of car, denoted by the subscript c, for any given household, i, is captured by the

following formula:

(1)

Where A, B, and C are constants, K represents the characteristics of the buyer, and

L represents the characteristics of the car. The constant A in Equation (Eqn.) 1 can be

related to the type of car that, in effect, represents a bias that a consumer might have

28

towards a vehicle. For instance, range anxiety might well make a consumer wary of an

EV and this may be captured in the constant A, driving down the probability of choosing

an electric car. Likewise, styling, or fuel economy might well have an influence, as could

environmental considerations.

Though Jeihani and Sibdari (2010) used only one standalone constant, A, one

could imagine including a great many such constants in an attempt to capture a greater

range of consumer biases. One could imagine developing an input to the car

characteristics that would include a national defense cost per vehicle, where that cost

reflects the investment made by the DoD to maintain shipping lanes, providing support to

governments that are critical for oil imports, etc. Those costs would then appear as a

price per vehicle, which could be denoted as a national defense cost, ND, that is weighted

by the fuel efficiency of a particular vehicle. Likewise, one could attempt to capture the

economic costs of the oil trade deficit to the nation, the approximately one billion dollar

daily deficit. Those costs could also appear as a price per vehicle and, denoted as a trade

imbalance cost, TI, would also be weighted by a vehicle’s fuel efficiency.

The constant B weights the characteristics of the buyer towards certain car types,

while the constant C weights the vehicle characteristics.

The constant Ki captures household characteristics including both employment

status and income while L captures the initial vehicle cost and the cost per mile for

vehicle locomotion.

B. DEVELOPMENT OF THE PREDICTIVE SALES MODEL

Rather than calculating the relative probability of choosing one car over another,

the model presented in this section solves for sales volume of a particular vehicle class.

In all, three vehicle classes are looked at, SUV, hybrid, and EV. Historical data on both

SUV and hybrid vehicles is used to establish coefficients that allow a least squares fit on

that model to that historical data. Where specific vehicle averages are necessary, for

example average sales price of the class of vehicle, a representative vehicle was chosen.

The Ford Explorer represents the SUV class of vehicles, and the Toyota Prius represents

the hybrid class of vehicles.

29

This approach allows the assignment of consumer coefficients from sales data and

presumes that these coefficients will remain the same for the foreseeable future. As there

is insufficient data for a meaningful historical analysis of EV sales, the coefficients

derived for hybrid vehicles are used as a proxy in predicting EV sales, assuming the two

are closely related. Those coefficients are then used in the model to predict future sales

of EVs.

Notably, Jeihani and Sibdari (2010) did not adjust for inflation when analyzing

historical data they collected. Adjusting the historical data used herein was considered;

however, the model is designed to reflect people’s choices at a fixed point in time. When

consumers go to purchase a vehicle they are considering what the price of gasoline

currently is, what their income currently is, etc. They may also be considering what they

expect the value of those items to be in the future, but they are concerned with absolute

amounts. The model shows fixed choices from moment to moment. If high levels of

inflation were anticipated going forward, that could influence consumer purchasing

decisions in general (e.g., buy assets rather than hold onto money), but none of the

purchase periods analyzed in this thesis took place during periods where inflation was

particularly high. From 1991 through 2010, no calendar year saw inflation greater than

4.2 percent (U.S. Department of Labor, 2011). In fact, adjusting for inflation would

flatten the changes in the economic data. Rather than reflecting reality, it would actually

skew the reality that the consumer faced at the time of their purchase decision.

Assuming a linear relationship between the change in the number of hybrid

vehicles sold in any year, ΔShy, and any of the influencing parameters, as was the basis

for the Jehani and Sibdari (2010) model, the number of hybrid vehicles sold in any year,

Shy, can be modeled and the model be used to project future sales. Future sales depends

on many economic and perception factors. Among the national economic parameters the

Jehani and Sibdari (2010) model included were: the median household income I, in the

modeled year, the unemployment level U, and the price of gasoline G. For the vehicular

economic parameters the model included the gas mileage, Mhy, as represented by the

number of miles the vehicle can travel per gallon of gasoline and the hybrid vehicle price

Phy. Finally, all parameters associated with customer perception such as comfort,

30

reliability, social appearance, environmental stewardship, etc. were lumped into a single

coefficient Ahy. In essence, the coefficient represents the total combination of factors the

consumer is considering that have not been specifically accounted for elsewhere in the

equation.

Another way to look at the effect of policy choices is by examining the fraction of

new sales, ΔSannual, in any year relative to the total sales Sannual. This fraction is the sales

strain, and is represented by:

(2)

A linear relationship between the sales strain and any influencing parameter is the

simplest approach to modeling the sales figures. For example, the effect of household

income, I, on the sales strain of hybrid vehicles when isolated from other parameters that

influence sales, would be:

(3)

Where ΔSIhy is the change in the sales figures of hybrid vehicles due to the change

in household income. As the overall sales figure Shy increases so does the change in that

figure due changes in any parameter such as I. The fundamental assumption in Eqn. (3)

is that ΔSIhy varies linearly both with I and Shy. The coefficient Bhy is a proportionality

factor to be derived empirically using historic data.

The effect of the price of gasoline, G, and vehicle mileage can be combined to

provide the cost per mile driven CPMhy where:

hy

hy MGCPM = (4)

Dependences similar to Eqn. 3 for the effects of U, CPMhy and Phy on the sales

figures can be derived to provide ΔSUhy, ΔSCPMhy

hy and ΔSPhyhy, which are the change in

the sales figures when each of the representative parameters is isolated from the others.

The proportionality coefficients to be assigned to these dependences are Chy, Dhy and Ehy

respectively. All three coefficients need to be determined empirically using historic data.

31

The combined effect of all these parameters on the annual sales figures of hybrid

vehicles can be obtained by superposition or a simple addition:

hyhyhyhyhyhyhyhy SPECPMDUCIBS )( +++=∆ (5)

Eqn. 5 can now be integrated for Shy, then adding Ahy, the integration coefficient,

which is described above, to provide:

)exp( hyhyhyhyhyhyhyhy PECPMDUCIBAS ++++= (6)

Notably, this approach results in an exponential relationship between sales

volume and income, unemployment, fuel costs and purchase price. Taking the natural

log of this equation yields:

hyhyhyhyhyhyhyhy PECPMDUCIBAS ++++=)ln( (7)

Note that Eqn. 7 is the immediate result of the integration of Eqn. 5. However,

Equation 6 was shown first to show the relationship between this derivation and the

model presented by Jehani and Sibdari (2010). This expresses the relationship between

the various parameters that are likely to influence the sales volume and the sales volume

itself. Actual numerical values can be determined after obtaining the values of the

coefficients Bhy, Chy, Dhy and Ehy, with an offset factor given by Ahy. To derive the five

empirical coefficients of Eqn. (7), one needs to obtain historical data concerning

household income, unemployment, price of gasoline, gas mileage and car prices for at

least five years. To the extent data is available, a similar equation can be derived for each

vehicle model or model group.

The equation representing the sales figures of SUV vehicles was derived similarly

to Eqn. (7) to yield:

SUVSUVSUVSUVSUVSUVSUVSUV PECPMDUCIBAS ++++=)ln( (8)

Again, the coefficients ASUV, BSUV, CSUV, DSUV, and ESUV are empirical

coefficients. Similar to those of the hybrid vehicle, they are to be derived using at least

five years of historical data.

Unlike SUVs and hybrid vehicles, the available historic data on the sales volume

of EVs is inconsistent with extremely small numbers. Consequently, those figures were

32

not considered reliable for the purpose of projecting future sales of EVs and the potential

response to market-changing policies. Instead, the coefficients derived for the hybrid

vehicle using historical data were retained as proxies for the EV sales volume model. It

is reasonable to presume that factors influencing people’s decision to purchase hybrid

vehicles will similarly influence their decision to purchase EVs. As with hybrids, it is

likely that the first consumers to purchase EVs do so for a myriad of concerns (e.g.,

environmental, status). As sales of hybrids grew, economic forces such as declining

vehicle price and increase in gasoline would have become the dominant market drivers.

The same is presumed to be true with EVs. Therefore, only one parameter from the

hybrid model Eqn. 7 was replaced. CPMhy was replaced with CPMEV, which was derived

using the representative cost of electricity per kWhr and the number of miles driven by

the EV per kWhr. Accordingly, the model describing the sales volume of electric vehicle

is represented by:

( ) ( ) ( ) ( ) ( )EVhyEVhyhyhyhyEV PEPMDUCIBAS ×+×+×+×+=ln (9)

The coefficients representing the sales volume of hybrids (Eqn. 7), and

consequently EVs, and SUVs (Eqn. 8) were derived using historic data. Table 1 contains

the historic data used, encompassing the years 1991 to 2010 for SUVs and 2000–2010 for

hybrids. More than five years of data was available for each vehicle group. With only

five years available, the five coefficients of each model (Eqns. 7 and 8) could be

determined by solving five algebraic equations. With additional years available, the

coefficients could be determined by fitting ln(Shy) and ln(SSUV), as determined by Eqns. 7

and 8 respectively, to the sales volumes as shown in Table 1, through adjustment of the

coefficients of these equations until the overlap between the historic data and the analytic

data is optimized.

Figures 4 and 5 graphically represent the outcome of the optimal fit between the

actual sales figures and the modeled figures. Optimization was achieved by minimization

of the root mean square of the differences between actual and modeled figures. Table 2

shows the coefficients of the two models as derived through this analysis.

33

As previously explained:

• Ahy = all consumer preferences and choice factors not specifically addressed elsewhere in the equation,

• Shy = the sales volume of hybrids,

• Bhy = weight factor for societal annual income,

• I = societal average annual income,

• Chy = weight factor for unemployment,

• U= total unemployment,

• Dhy = cost of locomotion factor,

• PMhy = price per mile for locomotion,

• Ehy = weight factor for price of hybrid,

• Phy = price of hybrid vehicle

Table 1. Historical Data for SUVs and Hybrid Vehicles

year income $ ^^^

Gas Price $/gallon **

SUV Sales ***

Hybrid Sales ^

Total Un-employment ^^

SUV Price $ †

Hybrid Price $ ††

Hybrid MPG †††

SUV MPG §

1991 30,126 1.098 1,095,000 8628000 15,747 16 1992 30,636 1.087 1,003,000 9613000 16,692 16 1993 31,241 1.067 1,311,000 8940000 17,550 17 1994 32,264 1.072 1,623,000 7996000 18,860 16 1995 34,076 1.103 1,816,000 7404000 22,305 16 1996 35,492 1.192 1,890,000 7236000 21,170 16 1997 37,005 1.189 2,450,000 6739000 21,485 16 1998 38,885 1.017 2,581,000 6210000 21,560 16 1999 40,696 1.116 2,831,000 5880000 22,070 16 2000 41,990 1.462 3,143,000 9,350 5692000 23,480 19,995 41 16 2001 42,228 1.384 3,450,000 20,282 6801000 25,210 19,995 41 17 2002 42,409 1.313 4,191,000 36,035 8378000 24,585 19,995 41 16 2003 43,318 1.516 4,118,000 47,600 8774000 26,285 19,995 41 15 2004 44,334 1.812 4,713,000 84,199 8149000 26,600 20,295 46 15 2005 46,326 2.24 4,084,000 209,711 7591000 27,165 21,275 46 16 2006 48,201 2.533 3,757,000 252,636 7001000 26,530 16,213* 46 16 2007 50,233 2.7 4,231,000 352,274 7078000 25,370 20,601* 46 16 2008 50,303 3.26 3,987,000 312,386 8924000 26,495 21,500 46 16 2009 49,777 2.36 2,296,000 290,271 14265000 28,470 22,000 46 16 2010 50,000 2.79 3,000,000 274,210 14825000 29,280 22,800 50 17 *Includes weighted average of rebate that was offered during some or all of the year.

34

Legend for Table 1:

** Gas price $/gallon: From Jeihani, 2010; U.S. Energy Information Administration; 2011b

*** SUV sales: From U.S. Department of Energy, n.d. (only applicable to years 1991–2009. Year 2010 is an estimate).

^ Hybrid sales: From U.S. Department of Energy, 2011c

^^ Total unemployment: From Bureau, n.d.

^^^ Income: From Census, 2009 (only applicable to years 1991–2009. Year 2010 is an estimate).

†† Hybrid price: From 2001 Toyota Prius, 2006; 2002 Toyota Prius, 2006; 2003 Toyota Prius, 2006; 2004 Toyota Prius, n.d.; 2005 Toyota Prius, 2006; 2006 Toyota Prius, n.d.; 2007 Toyota Prius, n.d.; 2008 Toyota Prius, n.d.; 2009 Toyota Prius, n.d.;

† SUV Price: From 1991 Ford Explorer, n.d.; 1992 Ford Explorer, n.d.; 1993 Ford Explorer, n.d.; 1994 Ford Explorer, n.d.; 1995 Ford Explorer, n.d.; 1996 Ford Explorer, n.d.; 1997 Ford Explorer, n.d.; 1998 Ford Explorer, n.d.; 1999 Ford Explorer, n.d.; 2000 Ford Explorer, n.d.; 2001 Ford Explorer, n.d.; 2002 Ford Explorer, n.d.; 2003 Ford Explorer, n.d.; 2004 Ford Explorer, n.d.; 2005 Ford Explorer, n.d.; 2006 Ford Explorer, n.d.; 2007 Ford Explorer, n.d.; 2008 Ford Explorer, n.d.; 2009 Ford Explorer, n.d.; 2010 Ford Explorer, n.d.

††† Hybrid MPG: After Find a Car, n.d.

§ SUV MPG: After Find a Car, n.d.

* Information to perform weighted average for rebate: After Federal Tax, n.d.

Figures 4 and 5 show that, after optimization of the coefficient, Eqns. 7 and 8

yielded generally good approximations of the actual sales data.

Although the match between the actual and modeled sales volume is good, its

application is limited. One should note that the two models (Eqns. 7 and 8) are

independent of each other (i.e., as modeled, the sales of one group does not affect the

sales of the other group (non-cannibalization)). Such independence is justified when the

sales volume of hybrids or EVs is small relative to the total vehicle market; however,

when the sales of one group overwhelm the market, the model must be reexamined. This

failure of the model would occur at some point as the market responds so favorably to

government policies that the use of internal combustion engines is almost completely

abandoned in favor of EVs. Lacking a detailed analysis of the points of failure, it is

assumed herein that the model should be used primarily as the predictor of trends (i.e.,

significant increase or decrease in the sales of certain vehicle groups, rather than the

predictor of actual sales volumes).

35

Figure 4. Comparison of the Variation of ln(Shy) with Year of Sale as Derived by Model

(blue) with Historic Data (red).

Figure 5. Comparison of the Variation of ln(SSUV) with Year of Sale as Derived by

Model (blue) with Historic Data (red)

The resulting coefficients are contained in Table 2.

Table 2. Coefficients of the Sales Volume Models (Eqns. 7 and 8) of Hybrid and SUV Vehicles as Derived by Fitting These Equations to Historical Sales Data (Table 1)

A B C D E Hybrid/electric 3.6 2.16*10–4 -1*10–13 31.1 -1.66*10–4

SUV 13.36 2.01*10–4 -1.4*10–7 -10.8 -1.97*10–4

36

Note that Dhy, which represents the response of the hybrid vehicle sales figures to

gas prices, is positive whereas DSUV is negative. This is intuitively plausible and

represents the expected fact that higher gas prices provide an incentive for the purchase

of hybrid or electric vehicles whereas they provide a disincentive to the purchase (or

ownership) of SUVs. This is the basis for the government policy to be proposed below.

C. APPLICATION OF THE MODEL TO FORECAST SALES

The hybrid vehicle sales volume model (Eqn. 7) was used to project future sales

volumes of EVs and the response to various government policies. To this end, the

various columns of Table 1 were filled using projected data (e.g., expectations of future

median household income or unemployment levels). The government policy choices, the

gasoline excise tax per gallon and the rebate to the buyers of EVs, are parameters that can

be adjusted by the user to achieve a desired result (e.g., rapid growth in the sales of EV or

rapid decline in the use of gasoline).

Although the model was tested using a wide variety of parameters, the following

parameters were ultimately selected as representing a good balance of the positive and

negative effects of available policy choices and conservative assumptions about future

economic conditions. The assumptions used herein are as follows:

• The gasoline excise tax is $2 per gallon in 2012, $4 the following year, and $5 in 2014 through 2018. (The use of funds collected through the tax, for example, offsetting the expense of the tax for low income individuals, is discussed in detail in Chapter V.)

• The EV price for 2011 uses the Manufacturer’s Suggested Retail Price (MSRP) of the entry-level 2012 model year Nissan Leaf, $35,200 (Nissan) and a three percent annual decrease in price thereafter. The expected MSRP does not include the government subsidy described below. This decrease is a conservative number considering the anticipated rapid change in volume, which will create economies of scale. However, this reflects a similar conservative determination made by Brooker, that the price of EV batteries will decrease by three percent per year (2010).

• Though soon after publication of this work, actual 2011 sales volumes may be available for EVs, the 2011 sales volume used here was obtained from the predictive model. The 2010 EV sales were estimated as 10 percent of the hybrid vehicles sales volume of 274,210 (DOE, 2011c).

37

• The 2011 total sales cost of the EV reduces the MSRP described above by the $7,500 tax rebate that is currently available for purchase of an all-electric vehicle. Beginning in 2012 and thereafter, the total cost of the EV begins with the MSRP price described above and further reduces it by $15,000, the amount of tax rebate used for this particular predictive analysis.

• Beginning in 2008 with an EV cost per mile of three cents (Electric, 2007), the model assumes a three percent annual increase in the price per mile for the EV. This presumes the cost of electricity will increase with increased demand, though this could be offset if batteries become more efficient at recharging or otherwise become more efficient.

• Gallons of gas sold per year begins with the actual gallons sold at retail in 2009, 18,176,124,000 gallons (EIA, n.d. a.), leaves the amount constant for 2010 and 2011, and assumes a five percent decrease in 2012, the first year a gasoline tax is imposed in this model. The five percent decrease assumes that imposition of the two dollar tax will induce a decrease in the demand for gasoline. Each year beyond 2012, the decrease in gasoline sales is determined by multiplying the number of EVs sold the prior year by 490 gallons. The average driver drives 13,476 miles per year (Federal Highway Administration, 2011). Current CAFE standards provide for an average vehicle gas mileage of 27.5 mpg (National Highway Traffic Safety Administration, n.d.). Therefore, multiplication of these two figures provides an estimate of the gallons of gasoline used by each vehicle per year. Assuming that each EV sold the previous year will reduce sales of gasoline-powered vehicles by one that equates to 490 fewer gallons of gasoline sold in the current year for each EV sold the prior year. In fact, the decrease is likely higher due to drivers who purchase EVs throughout the year and other factors, but this is the estimate used in the model. A further benefit of this assumption is that it indirectly introduces a coupling between the sales volume of EVs and sales of other vehicles. It assumes that sales of EVs will cannibalize sales of all other vehicles; however, this is an incomplete and very partial coupling that does not overcome the weakness of the model as indicated above.

• The unemployment factor begins in 2010 with the published total unemployment figure of 14,825,000 (Bureau of Labor Statistics, n.d.) and assumes a one percent reduction each year.

• The income factor uses the 2009 median annual income of $49,777 (Census Bureau, 2009) and assumes a one percent increase each year.

• The base price of gasoline, prior to the introduction of the excise tax, begins with an average price of three dollars per gallon in 2011 and

38

increases the price by five percent per year. That assumes that the world oil market will continue to see an increase in demand even as demand in the U.S. declines.

Applying the policy and economic assumptions above to Eqn. 7, taken together

with the coefficients from Table 2, provides the projections of sales of EV (Figure 6)

between 2011–2018.

Figure 6. Predicted EV Sales for the Period 2011–2018 in Response to Taxation and Subsidy Policies as Projected by the Model

Figure 6 shows that electric vehicle sales will begin rising rapidly in a “hockey

stick” manner once, the policies choices listed above are introduced in 2012. A rebate of

$15,000 to each EV buyer is expected to induce this “hockey stick” effect. In this

scenario, the source of the rebate is the revenue generated from the gasoline excise tax.

Table 3 shows the annual excise taxes predicted to be collected each year through

2018, the amounts to be distributed as rebates for EV purchases, and the remaining sums

to be used to implement other EV adoption policies (e.g., refunds to low-income families

and recharge infrastructure build out incentives). Note that until 2017, the incentive

rebates represent less than 25 percent of the collected tax.

39

Table 3. Excise Tax Collected

Year Total Gasoline Excise Tax Collected

Tax Used For EV Rebates

Gas Tax Net of EV Rebate

2012 $34,534,635,600 $3,982,118,284 $30,552,517,316 2013 $68,548,941,078 $5,438,617,192 $63,110,323,885 2014 $84,797,868,872 $7,405,192,521 $77,392,676,352 2015 $83,588,354,094 $10,054,056,083 $73,534,298,011 2016 $81,946,191,600 $13,613,986,738 $68,332,204,863 2017 $79,722,573,766 $18,388,609,526 $61,333,964,241 2018 $76,719,100,877 $24,780,556,145 $51,938,544,732

D. FORMULAIC SHORTCOMINGS

1. Use of Hybrid Coefficients

The coefficients for the EV model were transferred from the coefficients for the

hybrid model, as there is insufficient historical data to solve for the EV model

coefficients independently. However, in order to make that change, the cost of

locomotion for EVs was adjusted to reflect the cost of electricity used to recharge EVs,

versus the price of gasoline that influenced the cost of locomotion for the hybrid vehicle.

That means that the EV predictive model is somewhat untethered from the cost of

gasoline. The catchall coefficient may still capture the psychological influence of the

price of gasoline, but the economic consequences are no longer directly a part of the EV

predictive model. However, the economic effect of changes to the price of gasoline still

impacts the predicted sales for SUVs (Eqn. 8). If that model were run to predict future

sales, higher gasoline prices would result in fewer sales of SUVs, but, as constructed, the

EV model may not sufficiently account for the positive influence on EV sales attributable

to increases in the price of gasoline. The likely result is that the model under predicts the

increase in EV sales as the price of gasoline increases.

2. The Model Does not Reflect the Bounded Nature of the Market

A fundamental reality, reflected in most in depth economic analyses, is that most

markets are finite. Consequently, although the total vehicle market may experience year-

to-year growth, the total number of vehicles sold in any year is finite. Thus, if a new

make, model or entirely new type of vehicle is introduced, its sales volume generally

40

comes at the expense of another group or groups of vehicles. A detailed analysis that

intended to predict sales volumes with accuracy and precision at all sales levels must

account for this zero sum game. Unfortunately, such an analysis is exceptionally

complex. It is beyond the scope and purpose of this work to develop a model of that

complexity.

The EV predictive formula (Eqn. 9) does not assume a zero sum game. In other-

words, there is no limit on the total sales of vehicles of all classes combined. Increased

sales of EVs do not automatically result in the loss of sales for any other category of

vehicles. While the factors that tend to increase EV sales are likely to decrease other

vehicle sales if their predictive models were run with the same assumptions, there is a

less than perfect link. Fortunately, when the sales volume of any group is small, its effect

on the sales volume of other groups is negligibly small. Consequently, a model

projecting its variation under the effects of certain market drivers can be decoupled from

other groups without introducing an unduly large error. This was the underlying

assumption for the Jehani and Sibdari (2010) model and is the basis for the model used

herein.

The consequence of this assumption is that the model can be a good predictor of

the onset of certain trends (e.g., the “bend” of a “hockey stick” type growth), which

occurs while the sales volume is still small. It will fail to project the actual sales or their

impact on other segments of the market (e.g., sales of gasoline), once those sales have

grown to a level that is significant relative to the rest of the market. At the low end of the

changes in sales, that is probably not significant, but it reinforces the premise, especially

at higher volumes of sales, that the model is designed to demonstrate trends, movement

and changes from various policy choices. It will identify the “hockey stick” turning point

seen on the graph where the rate of sales dramatically increases. It is not intended to

predict actual specific sales numbers.

3. Inherent Risks of Prediction

As mentioned previously, past performance is no guarantee of future success.

The model is based on historical data of the effects of gasoline price increases and other

41

factors on consumer vehicle preference. Furthermore, there is no real world data to

review an increase in the price of gasoline of the magnitude suggested herein. As

gasoline prices reach high levels, other consequences will almost certainly emerge. For

example, consumers may attempt to move closer to work and walk or bike, foregoing a

motor vehicle all together. Though this would still diminish the nation’s overall use of

oil, it might also diminish EV sales.

42

THIS PAGE INTENTIONALLY LEFT BLANK

43

V. CONCLUSIONS

A. SUGGESTED COURSE OF ACTION

The modeling results demonstrate that it is possible to alter consumer behavior

significantly in favor of EV purchases by using an excise tax to raise the price of gasoline

and using those funds in part to bring down the cost of EVs. Each step in lowering the

use of gasoline by automobiles is a step towards energy independence and greater

national and homeland security. Therefore, in a program announced significantly ahead

of time, the federal government should implement an excise tax on retail gasoline

purchases. The funds raised should go directly towards a tax credit or point of purchase

credit for the purchase of EVs. The additional funds raised should be used to offset the

effect of the tax on the lowest income segment in the population. That concern is

addressed in more detail later in this chapter. Additionally, it is important to use those

funds as incentive to propel the creation of a recharging infrastructure in the United

States. The ultimate goal is an all EV automobile market and the end of the internal

combustion automobile engine, without diminishment of the U.S. driving experience.

1. Excise Tax on Retail Gasoline Purchases

A tax on the sale of retail gasoline should be implemented. As was demonstrated

by the model, an excise tax of between two and five dollars raises more funds than is

required for rebates, even where the rebate is $15,000, double the current rebate on new

EV purchases. In order to allow consumers to prepare for the economic consequences of

the excise tax, and with the goal of seeing change at the beginning of implementation, it

is important that the policy be announced ahead of time. It takes one to two years of

gasoline price increases before there are real shifts in purchase behaviors (Goodwin,

2004; Jeihani & Sibdari, 2010).

Phasing in the tax serves two purposes. First, though announcing the increase

ahead of time should allow for a more immediate response to the price increases, there

are inevitably consumers who cannot or will not react immediately. A phase in allows

some of those consumers to adjust before the full tax is in place. Additionally, a phase in

44

would allow any unanticipated consequences to occur in a more controlled manner,

diminishing their magnitude and allowing more time to respond appropriately.

Eventually, economies of scale should bring down the price of EVs so that they

are sufficiently competitive without government rebates. According to the DOE, just the

creation of battery manufacturing plants, spurred by recovery act matching funds, is

lowering battery prices through economies of scale (Stimulus, 2010). As demand

increases and production rises to meet the demand, prices should fall allowing market

forces to drive the price reductions.

2. Recharge Infrastructure

If EVs are to be adopted nationwide, a well-designed and widely distributed

network of charging stations is imperative (Electric, 2010). A portion of the funds

collected by the excise tax net of EV rebates should be used to assist in the creation of

this infrastructure. Column four of Table 3 illustrates the additional monies that will be

available to policymakers. A portion of those funds will be required to offset the

economic hardship the gasoline tax may create for low-income families. That issue is

discussed in more detail below.

Range anxiety is a phenomenon not present with hybrid vehicles and, therefore,

not captured by the coefficients developed from hybrid vehicle historical data. Range

anxiety is particularly troublesome because it is more than a reflection of the ability of a

driver to go from point A to point B. It reflects that a driver may not try to go from point

A to point B even when the EV is fully capable of traveling that distance. Bostford and

Szczepanek describe this phenomenon well (2009). They describe range anxiety using an

anecdote from the Tokyo Electric Power Company (TEPCO). In 2007, TEPCO

introduced electric service vehicles and tracked employee usage of the vehicles over an 8

X 15 km service area. Initially, overnight charging was the only option. A few months

into the program they realized that EV drivers were only covering a small portion of the

service area. TEPCO responded by adding a fast charge station that could be used any

time of the day. After its installation, EV drivers accessed the service area in a similar

manner to conventional vehicle drivers. Most interestingly, the fast charger was rarely

45

used (Bostford & Szczepanek, 2009). The point is that the fast charger was not necessary

to meet the actual needs of the EVs, it was necessary to meet the psychological needs of

the EV drivers. It was necessary to counter range anxiety. If range anxiety is not

addressed, it may have a significant deleterious effect on the accuracy of the EV model.

Fortunately, recent stimulus funds have established the efficacy of dollar for

dollar grants in establishing a recharge infrastructure. As described previously, with two

million dollars of stimulus funds Oregon has created a recharge infrastructure covering

all of Interstate 5 that runs through the state. The facilities are spaced so that a driver is

never more than 50 miles between recharge locations (Oregon, 2010). It is not expected

that funds will be required to incentivize all recharge stations. As EV sales increase,

market forces may lead to the development of the infrastructure as well. Already some

retail establishments have determined that it is in their interest to put in recharge parking

spaces (DeLong, 2011). Nevertheless, especially given the concerns of range anxiety, to

truly jump start U.S. energy independence, significant investment should be put into

creating a recharge infrastructure, paid for by the retail gasoline excise tax.

3. New Power Stations and Infrastructure

EVs use far more energy than may be obvious. In fact, adding an EV to a

neighborhood, will increase the demand for electricity to the same or greater extent of

adding a new house to the neighborhood. For example, the Tesla Roadster contains a 56

kWh battery (Roadster Innovations, n.d.) that allows the vehicle to go 245 miles on a

single charge (Roadster Features, n.d.). By comparison, consumption by residential

utility customers averaged 908 kWh per month (How, 2011), meaning that the average

house uses about 30 kWh per day. The energy currently provided by gasoline will have

to come from power stations. This means the U.S. will need additional stations but also

additional transmission infrastructure as the current infrastructure is not sufficient

(Electric, 2010; Lewis, 2010).

Fortunately, power companies are largely financially successful (Gilbert, 2010). If

the regulatory obstacles are diminished, private industry should take over to build the

generation and transmission capabilities to meet the demand created by EVs (Gold 2009);

46

however, it takes time to build power stations and the transmission infrastructure to

support it. If the regulatory obstacles are removed, and the government’s plan to

encourage EV adoption is transparent, unambiguous, communicated and publicized one

to two years before it goes into effect, that will allow market forces to begin responding

so they are prepared to meet the rise in demand.

B. CRITICAL SECONDARY CONCERNS

1. Public Support Financially

Will the U.S. citizenry accept a substantial tax on the sale of gasoline? There is a

historical basis to believe they might, if citizens accept the security implications that are

alleviated by the tax. A parallel can be drawn to economic events during and after

WWII. Between the purchase of government war bonds, and the substantial federal

income tax paid by Americans, federal revenues were raised to never before seen

amounts, “$98.3 billion by 1945, nearly half the war-swollen GDP” (Sparrow, 2008).

Though the government’s propaganda campaign centered on the ethical imperative to

counter the Axis tyranny, combined with the self-interest realized by investing in war

bonds, research has shown that the government missed the mark. The real reason citizens

accepted the fiscal hardships was not for abstract ideals, rather it was when people were

faced with helping someone they could identify with (Sparrow, 2008). The idealized all-

American GI, the boy next door, is who they were helping. By paying taxes and buying

bonds, people saw themselves putting a gun and bullets directly in the hands of a GI.

This allowed for a sort of indirect participation in the war itself (Sparrow, 2008).

A similar view of the same social/psychological phenomena has been referred to

as the post-tragedy opportunity bubble (Breckenridge & Moghaddam, 2011). In their

paper, Breckenridge and Moghaddam look at the psychological similarities between the

attacks of 9/11 and the attack on Pearl Harbor (2011). They describe the fleeting moment

of opportunity where the populace rallies around its leaders, trusts them more, and,

because they are looking for a specific way to help, can be directed in a manner not

usually possible (Breckenridge & Moghaddam, 2011).

47

In the first State of the Union Address after 9/11, President Bush called on

Americans to give at least two years over the course of their lifetimes to the service of

their neighbors and nation (Bush, 2002). Unfortunately, that message did not resonate

with Americans in a way likely to make them feel like part of the fight against those who

attacked us. The success of fiscal participation during WWII turned on personalizing the

response, allowing the citizenry to feel it was truly participating in defeating the great

evil. The urgent desire to participate was given outlet in a contemporaneous ability to aid

in the defeat of the enemy. The outlet was immediate and direct, not an ephemeral and

vague channeling of that desire to some general purpose, at some undetermined time in

the future.

In reviewing the post 9/11 response, Breckenridge and Moghaddam (2011) show

that the government’s failure to provide meaningful participation resulted in a failure to

capture the public’s long-term engagement and support. There was no mechanism for the

citizenry to help defeat the great evil. Furthermore, the opportunity to harness the public

sentiment is fleeting. Once the bubble pops, the opportunity is essentially lost

(Breckenridge & Moghaddam, 2011).

Taken together, these examples show an uphill, though not impossible, task of

moving the citizenry of the United States to accept the sacrifice of a significant tax on

gasoline. A simple and straightforward message needs to make the case that there is an

ongoing and real evil threatening the nation, caused by dependence on foreign oil. In

countering this threat, the link must be clear in the minds of the citizenry; money spent at

the pump is buying back the very guns and bombs that are killing Americans. Secondly,

a personal and clear image must be established between the money paid and the lives

saved. While paying the tax may not put a gun in the hand of a GI, it can be shown to

take a gun out of the hands of the enemy. No more traumatic head injuries, no more

amputees, no more sophisticated plots to attack America.

Right now the country may not be ready to view the threat of dependence on

foreign oil in the same concrete terms as the bombing of Pearl Harbor or the events of

9/11. If that is the case, the government should nonetheless begin building the

framework. With so many enemies funded by oil sales, it is unfortunately only a matter

48

of time before another tragic event disrupts the landscape of the United States. When that

happens, leadership should be ready to ask for the participation of the citizenry, to ask for

their shared sacrifice as a direct and meaningful way to participate in defeating those who

seek to do us harm.

2. Public Support Politically

By using EVs as the mechanism to end energy dependence, this thesis promotes a

solution that both the political right (energy dependence) and the political left (EVs) have

vocally supported in the past. Though this paper is solely focused on the detrimental

effects energy dependence has on U.S. security, proposing a solution that is agreeable to

both sides of the political spectrum should ease adoption by the political system that must

choose whether or not and how to implement the plan.

3. Source of Energy Storage Materials

The current technology of choice for powering EVs is the lithium ion battery

(Forero, 2010; Tahil, 2006). There are several problems with reliance on lithium for

energy storage. In the first instance, there is disagreement with just how much

economically available lithium there is in the world, with some experts saying supplies

are quite limited (Tahil, 2006; Meridian, 2008) and others saying there is quite a bit

available (Evans, 2008). More troubling for national and homeland security is the

location of the largest known deposits of lithium. They are not in the United States. The

Andes Mountains in South America, specifically the area where the borders of Chile,

Bolivia and Argentina meet, contain a large majority of the world’s usable lithium

(Forero, 2010; Meridian, 2008), with Bolivia containing the largest known deposits in the

world (Risen, 2010). Recent discoveries in Afghanistan suggest that it too may possess

significant deposits of lithium (Risen, 2010). Additionally, current battery technology

relies on magnets of a type that depends on rare earth metals like neodymium, 95 percent

of which are produced in China (Ramsey, 2011).

It defeats the goal of energy independence if the U.S. simply trades one energy

dependency for another. As has been said, “We know that Bolivia can become the Saudi

Arabia of Lithium” (Romero, 2009). Policymakers must be mindful of this potential

49

development, but there are reasons to believe it is avoidable. In the first instance, unlike

reliance on oil, where the resource is consumed with each trip, with EVs, locally

produced electricity is consumed with each trip where additional lithium is only required

when the battery is replaced or a new vehicle is purchased.

There are also other potential sources of lithium. The Institute of Ocean Energy

at Saga University in Japan has described the research being conducted by Japan and

South Korea to enable harvesting of the 230 billion tons of lithium present in seawater

(Institute, n.d.).

Finally, there is reason to believe that lithium may not be the one and only source

of energy for EV batteries. The history of the EV battery shows a progression every few

years to a different source material (Chan, 2004). EV batteries have gone from

predominantly lead acid to nickel metal hydride and now to lithium-ion (Chan, 2004).

Development is constantly progressing on a variety of alternatives like aluminum air

batteries. Some research has shown reason to believe that metal air batteries—where the

cathode of the battery is air—could provide up to 11 times the energy density of the best

lithium-ion batteries currently available (Hamilton, 2009). Variety should be strongly

encouraged, with appropriate nudges (Thaler, 2008) to help orient the market’s focus

towards resources available within the U.S.

C. UNINTENDED CONSEQUENCES AND SECONDARY CONCERNS

1. Fungible Nature of the Oil Market

As described earlier in this thesis, oil is part of a fungible world market

(Chairman, 2007). That means that reducing oil to levels where the United States is

capable of providing all its petroleum needs, does not necessarily mean that the oil used

in the United States will be 100 percent domestically produced. As it stands currently,

the United States exports two million barrels of oil per day (EIA, 2010a, p. 128). As U.S.

demand decreases with the roll out of EVs, the U.S. may begin to export more of its oil.

As a fungible product, domestic oil prices will not necessarily be protected from the

effects of international disruption of oil supplies. However, if the United States is

50

capable of supplying its own oil requirements, then with proper planning it can ensure a

relatively uninterrupted supply of oil regardless of international supply disruptions.

For example, in considering what first responders can do in the event of localized

fuel shortages, it has been suggested that they should enter into firm contracts for fuel

(Leotta, 2006). At present, firm contracts would not solve the problem for the country as

a whole. If the world supply were disrupted, domestic sources could not fulfill the

contracts, because there simply would not be enough fuel to go around. However, once

energy independence is achieved, firm contracts with domestic suppliers could all be

fulfilled. Those portions of the government or private sector who wish to hedge against

supply disruptions could enter into futures contracts, or other contractual arrangements to

ensure a given supply at a given price.

2. A Regressive Excise Tax

A tax is regressive when it causes lower-income families to pay a higher

percentage of their income to the tax (Internal Revenue Service, n.d.). By adding a cost

per gallon of gasoline, the financial impact could have a disproportionate impact on

lower-income families. A simple way to lessen the impact would be in the form of a tax

credit that is phased out over a particular income level. To avoid bureaucratic expenses,

a national or regional average of both the price of gasoline and of the average gallons

consumed can be used. The credit will not precisely match the expense incurred, but can

be sufficiently harmonized to minimize the harm to lower-income families.

For example, in 2009, there were 8.8 million families living below the poverty

line. For an idea of what that measures for a family of four, made up of two adults and

two children, the poverty line was $21,756 (DeNavas-Walt, 2010). For purposes of this

example, assume that each of those families had one vehicle. As explained in Chapter

IV, the average driver uses 490 gallons of gasoline a year. Assuming policymakers felt

that all families below the poverty line should receive a rebate then in a year where the

gasoline excise tax was two dollars, 8.8 million families would receive a rebate of $980.

That would be a total rebate to those families of $8.624 billion. If the rebate was phased

out incrementally for families above the poverty line, even assuming another $8.624

51

billion was returned to those families, Table 3 shows that there would still be $13.3

billion left to use to promote recharge infrastructure growth ($30.55 billion left after EV

rebates, minus $8.624 billion times two).

3. Risk Increasing Cost of Goods / Inflation

Recent gasoline price increases have caused a corresponding increase in the cost

of goods, and may diminish consumers’ savings (CBO, 2008; Republican, 2010). Rising

gasoline prices can contribute to higher transportation costs, thereby raising expenses at

all stages of production (Republican, 2010). The proposed course of action in this thesis

minimizes that threat in a number of ways. In the first instance, the excise fee is only

levied on motor fuel. This does not include diesel fuel. In 2009, in the U.S., there was

approximately 16,878,000 gallons of diesel fuel sold per day (EIA, n.d. b) compared to

approximately 49,798,000 gallons of retail unleaded gasoline sold per day (EIA, n.d. a).

Furthermore, the 49,798,000 gallons of retail gasoline subject to the excise tax is only

about one sixth of the motor gasoline sales each day. The other categories of motor

gasoline sold each day are DTW, rack, and bulk (EIA, n.d. a). Therefore, governments

and businesses that obtain their gasoline in bulk would not be subject to the excise tax

under the implementation proposed in this thesis.

4. Significant Drop in the Price of Oil

The program may also be a victim of its own success. With sufficient numbers of

EVs on the roads, the demand for gasoline will take a measurable decline. Should a

sufficient number of other nations follow a similar course, the worldwide demand for

gasoline may drop significantly, thereby reducing the price of oil. As the price of oil and

therefore gasoline drops, the effect of the program may also decline. This can be

remedied by determining an appropriate floor for the price of gasoline and automatically

increasing the excise fee to maintain that floor.

52

D. FOLLOW-ON RESEARCH

1. A More Complex Version of the Present Model

Any number of factors can be added to the model. As long as there is historical

data measuring a particular factor, limitless terms can be added to the equation and the

coefficients solved for anew, using a least squares mean approach. For example, it would

be possible to separate out gasoline price as its own variable rather than simply have it

wrapped up in vehicle fuel efficiency. As part of fuel efficiency, it is really an economic

view of the relationship of gasoline price to vehicle choice. To the extent there is a

psychological impact from gasoline price and to allow gasoline effects to be more

specifically accounted for in the model, gasoline price could be included as an

independent factor. Though this psychological factor should be included in the catchall

term for the hybrid, and, therefore for the electric, vehicles, it may be possible to capture

the effect more directly. If a more complex model were developed for hybrid vehicles,

and the terms included appeared logically connected to the sale of EVs, then the more

complex model could be used, adopting the hybrid vehicle coefficients, to predict future

EV sales.

2. A More Complex Modeling System

One of the earlier described limits on the predictive model used in this thesis is

that the model is unbounded. In other-words, it does not envision a world where vehicle

purchases are essentially a zero sum game; however, the real world is essentially

bounded. The vast majority of consumers cannot purchase an EV and a gas-powered

vehicle. In the main, when it is time to go new car shopping, buyers are in the market to

purchase a single vehicle. This break with the reality of the marketplace means that

predictions of sales, especially at large volumes, develop obvious inaccuracies. If run far

enough into the future, the model will eventually predict sales of EVs that exceeds any

reasonable expectation of total vehicle sales for that year. It may be possible to create a

model that incorporates the linear relationship concept presented here, but that does so in

a way that bounds the universe of total vehicle sales.

53

3. Determine Coefficients for EVs Independently

As time marches forwarded and EV sales become a reality of the marketplace, it

may be possible to solve for the coefficients in the EV formula independently, rather than

borrowing from the hybrid equation. In the present model, there are five coefficients.

That means that there must be a minimum of five years of data in order to solve for the

coefficients. Once it is possible to do so, it should lead to an even more accurate EV

model.

4. Determining the Cost of Energy Dependence

Though the ramifications of energy dependence were discussed at great length,

this thesis did not attempt to put a dollar figure on the nation’s reliance on foreign oil.

Certainly the oil trade deficit is described in dollars, but the actual cost of securing the

flow of oil and of other policy choices the nation makes due to its dependence have not

been determined. Doing so may provide additional understanding to support policy

choices. If the cost were determined and then averaged over the various uses of that oil,

the true cost of oil being consumed could be established.

5. Infrastructure Questions

Though it is clear that additional recharge infrastructure is needed, and this thesis

suggests that monies raised from the excise tax should be used to spur the creation of the

infrastructure, further research would be helpful. It would be useful to determine the

ideal recharge locations to achieve the best bang for the buck. It would be useful to study

range anxiety so the recharge infrastructure met the psychological needs of consumers,

not just the physical needs of the EVs.

Similarly, it is clear that additional power generation will be required to meet the

increased demands on electricity from the growth of the EV market. Precisely how much

and where remains to be determined. Some of the demand may be met without

increasing actual power plant production. For example, smart meters that encourage non-

54

peak time charging can be installed. Some households might recharge using their own

solar or wind micro power generators. Whatever the need, it will be better met if it is

studied and well planned, rather than done haphazardly.

E. CONCLUSION

For 40 years, every president of the United States has proclaimed the critical

importance of energy independence. Time and again, the chains of foreign oil have

shackled the decisions of American officials; yet, nothing has been done. No pragmatic,

realistic step-by-step plan has been pursued to end this scourge on the American people.

This thesis proscribes the means whereby America can break free of those artificially

imposed restraints. There is a cost to achieving energy independence, and as shown

herein, that cost is two to five dollars on each gallon of retail gasoline sold. With

conviction, determination and selfless leadership, the United States can achieve energy

independence in a few short years.

55

LIST OF REFERENCES

1991 Ford Explorer. (n.d.). Retrieved November 20, 2011, from http://www.autotrader.com/1991-Ford-Explorer.jsp?modelId=11378

1992 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1992-Ford-Explorer.jsp?modelId=11378 1993 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1993-Ford-Explorer.jsp?modelId=11378 1994 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1994-Ford-Explorer.jsp?modelId=11378 1995 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1995-Ford-Explorer.jsp?modelId=11378 1996 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1996-Ford-Explorer.jsp?modelId=11378 1997 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1997-Ford-Explorer.jsp?modelId=11378 1998 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1998-Ford-Explorer.jsp?modelId=11378 1999 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/1999-Ford-Explorer.jsp?modelId=11378 2000 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2000-Ford-Explorer.jsp?modelId=11378 2001 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2001-Ford-Explorer.jsp?modelId=11378 2001 Toyota Prius. (2006, January). Retrieved November 20, 2011, from

http://www.epinions.com/specs/auto_Make-2001_Toyota_Prius 2002 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2002-Ford-Explorer.jsp?modelId=11378 2002 Toyota Prius. (2006, January). Retrieved November 20, 2011, from

http://www.epinions.com/specs/auto_Make-2001_Toyota_Prius

56

2003 Ford Explorer. (n.d.). Retrieved November 20, 2011, from http://www.autotrader.com/2003-Ford-Explorer.jsp?modelId=11378

2003 Toyota Prius. (2006, January). Retrieved November 20, 2011, from

http://www.epinions.com/specs/auto_Make-2001_Toyota_Prius 2004 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2004-Ford-Explorer.jsp?modelId=11378 2004 Toyota Prius. (n.d.). Retrieved November 20, 2011, from

http://consumerguideauto.howstuffworks.com/2004-toyota-prius.htm 2005 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2005-Ford-Explorer.jsp?modelId=11378 2005 Toyota Prius. (2006, January). Retrieved November 20, 2011, from

http://www.epinions.com/specs/auto_Make-2005_Toyota_Prius 2006 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2006-Ford-Explorer.jsp?modelId=11378 2006 Toyota Prius used cars for sale. (n.d.). Retrieved November 20, 2011, from

http://www.web2carz.com/toyota/prius/2006 2007 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2007-Ford-Explorer.jsp?modelId=11378 2007 Toyota Prius used cars for sale. (n.d.). Retrieved November 20, 2011, from

http://www.web2carz.com/toyota/prius/2007 2008 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2008-Ford-Explorer.jsp?modelId=11378 2008 Toyota Prius used cars for sale. (n.d.). Retrieved November 20, 2011, from

http://www.web2carz.com/toyota/prius/2008 2009 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2009-Ford-Explorer.jsp?modelId=11378 2009 Toyota Prius used cars for sale. (n.d.). Retrieved November 20, 2011, from

http://www.web2carz.com/toyota/prius/2009 2010 Ford Explorer. (n.d.). Retrieved November 20, 2011, from

http://www.autotrader.com/2010-Ford-Explorer.jsp?modelId=11378

57

2010 Toyota Prius used cars for sale. (n.d.). Retrieved November 20, 2011, from http://www.web2carz.com/toyota/prius/2010

Bedsworth, L. & Taylor, M. (2007). Learning from California’s zero-emission vehicle

program. California Economic Policy, 3(4). Berman, I. (2006, Jul 25). The economics of confronting Iran, Statement before the Joint

Economic Committee of the United States Congress. Retrieved September 20, 2010, from www.afpc.org/files/getContentPostAttachment/24

Botsford, C. & Szczepanek, A. (2009, May 13–16). Fast charging vs. slow charging:

pros and cons for the new age of electric vehicles. Presented at EVS24 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Stavanger, Norway.

Boyd, J. (1996). Notice of Public Hearing to consider amendments to the zero-emission

vehicle requirements for passenger cars and light-duty trucks. Retrieved September 9, 2011, from http://www.arb.ca.gov/board/books/010628/01–5–1.pdf

Breckenridge, J. N. & Moghaddam F. L. (September 2011). The post-tragedy

‘opportunity-bubble’ and the prospect of citizen engagement [The 9/11 Essays]. Homeland Security Affairs, 7. Retrieved September 9, 2011, from http://www.hsaj.org/?article=7.2.2

Broder, J. (2011, March 30). Obama sets goal of one-third cut in oil imports. New York

Times [on-line]. Retrieved March 31, 2011, from http://www.nytimes.com/2011/03/31/science/earth/31energy.html

Brooker, A., Thornton, M., & Rugh, J. (2010, January). Technology improvement

pathways to cost-effective vehicle electrification. National Renewable Energy Laboratory.

Bureau of Labor Statistics. (n.d.). Household data annual averages. Retrieved July 8,

2011, from http://www.bls.gov/cps/cpsaat1.pdf Bush, G. (2002, January, 29). State of the Union Address. Retrieved October 23, 2011,

from http://articles.cnn.com/2002–01–29/politics/bush.speech.txt_1_firefighter-returns-terrorist-training-camps-interim-leader?_s=PM:ALLPOLITICS

Calef, D. & Goble, R. (2007). The allure of technology: How France and California

promoted electric and hybrid vehicles to reduce urban air pollution. Policy Sci, 40, 1–34.

58

Census Bureau. (2009). Median household income by state: 1984 to 2009. Retrieved October 5, 2011, from http://www.census.gov/hhes/www/income/data/historical/household/H08_2009.xls

Chairman Lantos and ranking Republican Ros-Lehtinen, House of Representatives

Committee on Foreign Affairs, Foreign Policy and National Security Implications of Oil Dependence, (2007, March 22). Retrieved August 1, 2010 from http://internationalrelations.house.gov/110/34241.pdf

Chan, C.C. & Wong, Y.S. (2004, November/December). Electric vehicles charge

forward. IEEE Power and Energy Magazine. CNN Wire Staff. (2010, July 25). Chavez threatens to cut off oil to U.S. Retrieved August

1, 2010, from http://www.cnn.com/2010/WORLD/americas/07/25/venezuela.chavez.us/index.html

Cohen, A. (2007, March 2007). The national security consequences of oil dependency.

Heritage Lectures. Retrieved August 1, 2010, from www.heritage.org/research/nationalsecurity/hl1021.cfm

Congressional Budget Office. (2008). Effects of gasoline prices on driving behavior and

vehicle markets. Washington, DC; author. Retrieved December 24, 2010 from http://www.cbo.gov/ftpdocs/88xx/doc8893/01–14-GasolinePrices.pdf

Cote´, J. (2010, October 27). Electric cars to get boost from Bay Area grants. San

Francisco Chronicle [on-line]. Retrieved October 27, 2010, from http://articles.sfgate.com/2010–10–27/bay-area/24223494_1_nissan-leaf-electric-cars-chevrolet-volt

Cowden, D. (2008, summer). A brief inquiry into one of the technologies to achieve U.S.

energy security. Monterey, CA: Naval Postgraduate School. Daigneau, E. (2011, June). Planning for plug-ins. Governing. DeLong, J. (2011, January 18). Retail outlets adding electric car charging stations. USA

Today. Retrieved October 5, 2011, from http://www.usatoday.com/tech/news/2011–01–14-electric-car-charging-stations_N.htm

DeNavas-Walt, C., Proctor, B., Smith, J. (2010, September). Income, Poverty, and Health

Insurance Coverage in the United States: 2009. Retrieved October 5, 2011 from http://www.census.gov/prod/2010pubs/p60–238.pdf

59

Dickerman, L. & Harrison, J. (2010, March/April). A new car, a new grid. IEEE Power & Energy Magazine.

Electrical Construction & Maintenance. (2010, September). Electric vehicles. Retrieved

December 27, 2010, from http://ecmweb.com/market_trends/electric-vehicle-questions-answered-20100901/

Espey, M. (1996). Explaining the variation in elasticity estimates of gasoline demand in

the United States. Energy Journal, 17(3), 49. Etezadi-Amoli, M. (2010, July). Rapid-charge electric-vehicle stations. IEEE

Transactions on Power Deliver, 25(3). Electric Vehicle News. (2010, November 15). German driver logs 65,000 km in one year

in his Tesla Roadster. Retrieved November 15, 2010, from http://electric-vehicles-cars-bikes.blogspot.com/2010/11/german-driver-logs-65000-km-in-one-year.html

Electric Vehicles of Washington. (2007). EV conversions. Retrieved October 4, 2011,

from http://www.electricvehicleswa.com/electric-vehicle-hybrid-conversions.php Federal Emergency Management Agency. (2004). Lessons learned information sharing,

Department of Homeland Security: Continuity of operations: coordinating backup power and fuel supply. Washington, DC: author.

Federal Highway Administration. (2011, April 4). Average annual miles per driver by

age group. Retrieved October 3, 2011, from http://www.fhwa.dot.gov/ohim/onh00/bar8.htm

Federal tax credits for hybrids. (n.d.). Retrieved November 20, 2011, from

http://www.fueleconomy.gov/feg/tax_hybrid.shtml Find a Car. (n.d.). Retrieved November 20, 2011, from

http://www.fueleconomy.gov/feg/findacar.htm Forero, J. (2010, December 17). S. American mountains hold key to electric car’s future:

lithium for batteries. Washington Post. Retrieved December 20, 2010, from http://www.washingtonpost.com/wp-dyn/content/article/2010/12/17/AR2010121703351_pf.html

Federal Trade Commission. (2005). Gasoline price changes: The dynamic of supply,

demand, and competition. Retrieved December 24, 2010, from http://www.ftc.gov/reports/gasprices05/050705gaspricesrpt.pdf

60

Gilbert, J. (2010, September 13). Most profitable utilities by recurring EBITDA margin. Retrieved January 17, 2011, from http://www.snl.com/InteractiveX/ArticleAbstract.aspx?id=11690702

Gold, J. (2009, April 27). A modern electric grid: The new highway system? National Public Radio. Retrieved December 24, 2010, from http://www.npr.org/templates/story/story.php?storyId=103349614

Goodwin, P., Dargay, J. & Hanly, M. (2004, May). Elasticities of road traffic and fuel

consumption with respect to price and income: A review. Transport Reviews, 24(3), 275–292.

Gore, A. (1992). Earth in balance: Ecology and the human spirit. Boston, MA:

Houghton Mifflin. Haldis, P. & Franco, J. (2009). Analysts question financial attractiveness of PHEVs,

electric vehicles. Global Refining and Fuels Today, 1(77), 15. Hamilton, T. (2009, November 5). Betting on a metal-air battery breakthrough.

Technology Review. Retrieved November 14, 2010, from http://www.technologyreview.com/printer_friendly_article.aspx?id=23877

Hauser, R. (2007). Corn-based ethanol in Illinois and the U.S.: A report from the

Department of Agricultural and Consumer Economics. Champaign, IL: Department of Agricultural and Consumer Economics, University of Illinois.

Hammerschmidt, C. (2009, September 22). E-car battery price on rapid declining curve,

expert predicts. Retrieved July 10, 2011, from http://www.eetimes.com/electronics-news/4084603/E-car-battery-price-on-rapid-declining-curve-expert-predicts

Hensley, R., Knupfer, S., & Pinner, D. (2009, June). Electrifying cars: How three

industries will evolve. McKinsey Quarterly. Institute of Ocean Energy Saga University. (n.d.). Lithium—Recovery from seawater.

Retrieved July 13, 2011, from http://www.ioes.saga-u.ac.jp/ioes-study/li/recovery/seawater.html

Internal Revenue Service. (n.d.). Whys of taxes theme 3: Fairness in taxes lesson 2:

Regressive Taxes. Retrieved July 8, 2011, from http://www.irs.gov/app/understandingTaxes/student/whys_thm03_les02.jsp

International Association of Fire Chiefs. (2003). Performance of the fire service during the 2003 northeast blackout and the implications for critical infrastructure protection. Fairfax, VA: author.

61

Jeihani, M. & Sibdari, S. (2010). The impact of gas price trends on vehicle type choice. Journal of Economics and Economic Education Research, 11(2).

Kayser, H. (2000). Gasoline demand and car choice: estimating gasoline demand using

household information. Energy Economics, 22, 3331–3348. Kimery, A. (2008, August 18). Oil disruptions threaten national security. The Kimery

Report Blog. Retrieved April 11, 2011, from http://www.hstoday.us/blogs/the-kimery-report/blog/oil-disruptions-threaten-national-security/94c2f1e00a48f4dadf2cfb9734f72e88.html

Lee, J. (2010, July 15). Electric vehicles, advanced batteries, and American jobs:

Another piece of the puzzle in Holland, Michigan. White House Blog. Retrieved July 5, 2011, from http://www.whitehouse.gov/blog/2010/07/15/electric-vehicles-advanced-batteries-and-american-jobs-another-piece-puzzle-holland-

Lewis, T. (2010, February 23). How to create a resilient infrastructure in 20 years for 1 trillion dollars, create millions of jobs, transition to green transportation, and do all of this at no cost to government. Homeland Security Watch. Retrieved October 5, 2011, from http://www.hlswatch.com/2010/02/23/how-to-create-a-resilient-infrastructure-in-20-years-for-1-trillion-dollars-create-millions-of-jobs-transition-to-green-transportation-and-do-all-of-this-at-no-cost-to-government/

Leotta, K. (2006). Fuel price, availability, and mobility: What we can learn from North

Carolina in the aftermath of Hurricane Katrina, and oil shocks of the 1970s and early 1980s. Seattle, WA: Parsons Brinkerhoff.

Light, A. (1976, winter). Federalism and the energy crisis: A view from the States.

Journal of Federalism. McFarlane, R. & Woolsey, R. J. (2011, September 21). How to weaken the power of

foreign oil. New York Times, A31. Meridian International Research. (2008). Trouble with lithium 2. Retrieved November

14, 2010, from www.meridian-int-res.com/Projects/Lithium_Microscope.pdf Morris, M. (2007, March 28). Short-run motor gasoline demand model. Presented at

EIA Modeling and Data Conference, Washington, D.C. Retrieved December 20, 2010, from www.eia.gov/oiaf/aeo/conf/pdf/morris.ppt

National Highway Traffic Safety Administration. (n.d.). CAFE Overview—Frequently

asked questions. Retrieved October 5, 2011, from http://www.nhtsa.gov/cars/rules/CAFE/overview.htm

62

Nissan Leaf packages and features. (n.d.). Retrieved October 1, 2011, from http://www.nissanusa.com/leaf-electric-car/feature/pricing_information#/leaf-electric-car/feature/pricing_information

Paust, J. & Blaustein, A. (1974). The Arab oil weapon: A threat to International Peace.

American Journal of International Law, 68(410). Peckham, J. (2009). Electric cars seen taking 64% of U.S. car market by 2030. Diesel

Fuel News, 13(15), 25. Powers, J. (2010). Oil addiction: Fueling our enemies. Washington, DC: Truman

Project. Ramsey, M. (2011, February 14). Toyota tries to break reliance on China. Wall Street

Journal. Retrieved July 13, 2011, from http://online.wsj.com/article/SB10001424052748703583404576080213245888864.html

Republican Staff Commentary. (2010, December 16). Federal gasoline tax targeted for

new revenues Retrieved December 24, 2010, from http://jec.senate.gov/republicans/public/?a=Files.Serve&File_id=56428350-ebb5–48de-9b15-e7aa54d11588

Risen, James U.S. Identifies Vast Mineral Riches in Afghanistan New York Times on-

line (June 13, 2010) available at http://www.nytimes.com/2010/06/14/world/asia/14minerals.html (accessed on 11/14/2010)

Roadster/features and specs. (n.d.). Retrieved July 13, 2011, from

http://www.teslamotors.com/roadster/specs Roadster innovations/battery. (n.d.). Retrieved July 13, 2011, from

http://www.teslamotors.com/roadster/technology/battery Romero, S. (2009, February 3). In Bolivia: Untapped bounty meets nationalism. New

York Times. Retrieved November 14, 2010, from http://www.nytimes.com/2009/02/03/world/americas/03lithium.html

Sanger, D. & Landler, M. (2011, October 16). To isolate Iran: U.S. presses inspectors on

nuclear data. New York Times, A1.

63

Sherman, D. (2011, January). Electric propulsion: Clean, green, high-efficiency machines are finally ready for the road. Automobile. Retrieved March 7, 2011, from http://www.automobilemag.com/features/awards/1101_2011_technology_of_the_year_electric_propulsion/index.html

Sparrow, J. (2008). Buying our boys back: The mass foundations of fiscal citizenship in

World War II. Journal of Policy History, 20(2). Stewart, J. (2010, June 6). An energy independent future. The Dailey Show. Retrieved

November 14, 2010, from, http://www.thedailyshow.com/watch/wed-june-16–2010/an-energy-independent-future

Sweet, C. Not just a lot of hot air (2010, September 13). Wall Street Journal. Retrieved

November 7, 2010, from http://online.wsj.com/article/SB10001424052748704133804575198362283533710.html

Stimulus funds charged U.S. battery industry. (2010, July 20). Physics Today. Retrieved

March 18, 2011, from http://blogs.physicstoday.org/politics/2010/07/stimulus-funds-charged-us-batt.html

Tesla builds its cars around the driver. (n.d.). Retrieved December 26, 2010, from

http://www.teslamotors.com/goelectric Tahil, W. (2007, January). The trouble with lithium, implications of future PHEV

production for lithium demand. Retrieved November 14, 2010, from http://www.meridian-int-res.com/Projects/Lithium_Problem_2.pdf

Thaler, R. & Sunstein, C. (2008). Nudge: Improving decisions about health, wealth and

happiness. New Haven, CT: Yale University Press. Thomas, S. & Kerner, D. (2010, August). Defense energy resilience: Lessons from

ecology. The Letort Papers. Retrieved December 24, 2010, from http://www.strategicstudiesinstitute.army.mil/pubs/download.cfm?q=1011

Totty, M. (2010, February 22). The long road to an alternative-energy future. The Wall

Street Journal. Retrieved November 7, 2010, from http://online.wsj.com/article/SB10001424052748704757904575077312569582760.html

U.S. Census Bureau. (2011). Trade in goods with China. Retrieved October 2, 2011,

from http://www.census.gov/foreign-trade/balance/c5700.html

64

U.S. Department of Energy. (n.d.). New Light-Duty Vehicles Sold in U.S. Retrieved July 10, 2011, from www.afdc.energy.gov/afdc/data/docs/ldv_sales.xls

U.S. Department of Energy. (2011a). 2011–12 electric vehicles. Retrieved October 1,

2011, from http://www.fueleconomy.gov/feg/evsbs.shtml U. S. Department of Energy. (2011b, February 8). DOE releases new analysis showing

significant advances in electric vehicle deployment. Retrieved March 18, 2011, from http://www.energy.gov/news/10057.htm

U.S. Department of Energy. (2011c, September 30). Hybrid electric vehicle (HEV) sales

by model. Retrieved October 4, 2011, from http://www.afdc.energy.gov/afdc/data/vehicles.html#afv_hev

U.S. Department of Labor. (2011). Consumer price index, all urban consumers.

Retrieved October 17, 2011, from ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt

U.S. Energy Information Administration. (n.d. a). Refiner motor gasoline sales volume.

Retrieved July 8, 2011, from http://www.eia.gov/dnav/pet/pet_cons_refmg_d_nus_VTR_mgalpd_a.htm

U.S. Energy Information Administration. (n.d. b). Refiner sales volumes of other

petroleum products. Retrieved July 8, 2011, from http://www.eia.gov/dnav/pet/pet_cons_refoth_d_nus_VTR_mgalpd_a.htm

U.S. Energy Information Administration. (2002). Petroleum chronology of events 1970–

2000. Retrieved May 15, 2010, from ftp://ftp.eia.doe.gov/pub/oil_gas/petroleum/analysis_publications/chronology/petroleumchronology2000.htm

U.S. Energy Information Administration. (2008), Imported Oil as a Percent of Total

U.S. Consumption. Retrieved May 15, 2011, from http://www.eia.doe.gov/emeu/25opec/sld002.htm

U.S. Energy Information Administration. (2009). Use of oil: Energy explained, your guide to understanding energy. Retrieved December 24, 2010, from http://www.eia.gov/energyexplained/index.cfm?page=oil_use

U.S. Energy Information Administration. (2010a, August). Annual energy review 2009.

Retrieved October 2, 2011, from http://www.eia.gov/totalenergy/data/annual/pdf/aer.pdf

65

U.S. Energy Information Administration. (2010b). How dependent are we on foreign oil? Retrieved December 20, 2010, from http://tonto.eia.doe.gov/energy_in_brief/foreign_oil_dependence.cfm

U.S. Energy Information Administration. (2011a, February 25). Crude oil and total

petroleum imports top 15 countries. Retrieved March 7, 2011, from http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/company_level_imports/current/import.html

U.S. Energy Information Administration. (2011b). Data 1: All grades conventional EIA

[Worksheet]. Retrieved October 6, 2011, from http://www.eia.gov/oog/ftparea/wogirs/xls/pswrgvwag.xls

U.S. Energy Information Administration. (2011c, June 1). How much electricity does an

American home use? Retrieved July 13, 2011, from http://www.eia.gov/tools/faqs/faq.cfm?id=97&t=3

U.S. Energy Information Administration. (2011d, June). Oil imports and exports.

Retrieved October 2, 2011, from http://www.eia.gov/energyexplained/index.cfm?page=oil_imports

U.S. Energy Information Administration. (2011e). Oil: Crude and petroleum products

explained use of oil. Retrieved July 4, 2011, from http://www.eia.gov/energyexplained/index.cfm?page=oil_use

Whoriskey, P. (2009, October 16). The hydrogen car gets its fuel back. Washington

Post. Retrieved November 11, 2011, from http://www.washingtonpost.com/wp-dyn/content/article/2009/10/16/AR2009101601002_pf.html

Worth, R. (2010, August, 6). Tanker damage caused by attack: Inquiry finds. Retrieved

October 23, 2011, from http://www.nytimes.com/2010/08/07/world/middleeast/07tanker.html

66

THIS PAGE INTENTIONALLY LEFT BLANK

67

INITIAL DISTRIBUTION LIST

1. Defense Technical Information Center Ft. Belvoir, VA

2. Dudley Knox Library Naval Postgraduate School Monterey, CA

3. Francine Kerner Transportation Security Administration Arlington, VA

4. Gabriel Laufer University of Virginia Charlottesville, VA